General Motors Corporation

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

IN THE MATTER OF GENERAL MOTORS CORPORATION DISMISSAL ORDER, ETC.) IN REGARD TO ALLEGED VIOLATION OF SEC. 5 OF THE FEDERAL TRADE COMMISSION ACT Docket 9077. Complaint, March 1976-Final Order, June, 1982 On June 25, 1982, the Commission dismissed its complaint charging a major car manufacturer with violating Section 5 of the Federal Trade Commission Act by its use of a selective distribution system for new crash parts for GM automobiles and light trucks.

Appearances For the Commission: Donald K. Tenney, Alan H. Melnicoe and Myron L. Dale.

For the respondent: Edwin S. Rockefeller, Alan M. Frey, Richard Haddad and Charles L. Duffney, Bierbower Rockefeller Washington, D. , and Francis H. Dunne in-house counsel, Detroit, Mich., for Jerry S. Cohen, Kohnrespondent General Motors Corporation. Milstein Cohen Washington, D. , for intervenor National Automobile Dealers Association. Jonathan T. Howe, Jenner Block Chicago, Ill. and Donald A. Randall in-house counsel, Washington , for intervenor Automotive Service Councils, Inc. COMPLAINT The Federal Trade Commission, having reason to believe that General Motors Corporation, a Delaware corporation, has engaged in unfair methods of competition and unfair acts or practices in connection with the distribution of new service crash parts applicable to automobiles and light trucks assembled by General Motors Corporation, in violation of Section 5 of the Federal Trade Commission Act, as amended (15 D. C. 45) and that a proceeding in respect thereof would be in the public interest, issues its complaint, charging as follows:

1. For the purpose of this complaint, the following definitions shall apply:

(a) Automobiles are self-propelled, four-wheeled vehicles primarily for the transport of persons-they travel primarily on roads or streets and their seating capacity is for no more than 10 persons. (h) Light trucks are self-propelled vehicles, other than automo- (g) 464. Complaint biles, designed to carry a load or freight, having a gross vehicular weight of less than 10,000 pounds, and traveling primarily on roads or streets.

(c) Service Parts or Replacement Parts are new parts used to replace parts assembled as original equipment (DE Parts) in new automobiles and light trucks or used to replace servce parts previously installed thereon.

(d) Crash Parts refers to anyone or all of the following products: fenders, griles, bumpers, hoods, deck lids, doors, quarter panels, rear end panels, rocker panels, lamp assemblies, wheel opening panels fender and rear end caps, tail gates, radiator supports and shrouds and mouldings, (2Jincluding inner and outer panels and all components of these products as well as all parts necessary to attach the aforesaid to the bodies of automobiles and light trucks. (e) Service GM Crash Parts are service crash parts applicable to automobiles and light trucks assembled by General Motors Corporation, sometimes hereinafter referred to as the relevant parts. (f) Distribution refers to the business of distributors. Distributors are firms which either manufacture service crash parts or contract for their manufacture for the purpose of reselling them, principally to franchisees who wholesale or install the parts. Wholesalers are firms which resell service crash parts to installers but which may also install such parts. They neither manufacture service crash parts nor do they contract for their manufacture.

2. Respondent General Motors Corporation (hereinafter "GM") is and at all times relevant herein has been a Delaware corporation; its headquarters are at 3044 W. Grand Boulevard, Detroit, Michigan. 3. GM is now and for many years has been engaged in the manufacture, sale and distribution of a wide variety of products including automobiles, trucks, buses, diesel locomotives, diesel engines, earth moving equipment, household appliances and automotive parts.

4. In 1972, GM had sales of $30.4 bilion, net earnings after taxes of $2.16 bilion, and total assets as of December 3I, 1972 of $18.3 bilion, ranking first in sales and profits and second in assets among the nation s largest industrial corporations. In 1975, GM had sales of $35.7 bilion and net earnings after taxes of $1.3 bilion. 5. GM is the largest manufacturer of automobiles and light trucks in the United States. Its principal domestic lines include Chevrolet, Pontiac, Oldsmobile, Buick and Cadilac automobiles and light trucks. In 1972, its total domestic sales of automobiles alone Complaint 99 F.

amounted to 4 823 827 units, 43% of the U.S. market and 52% of S. sales by domestic manufacturers. (3) 6. GM sells and for some time past has sold substantial amounts of crash parts. In 1972, GM's sales of service GM crash parts exceeded $250 milion.

7. In the course and conduct of its business, respondent GM is and for some time past has been engaged in selling service GM crash parts throughout various States of the United States, and has caused such part to be shipped to purchasers in various other states. In so doing, GM is and at all times relevant herein has been engaged in a continuous and substantial course of trade in commerce and has affected commerce as ucommerce" is defined in the amended Federal Trade Commission Act.

8. The number of automobile and light truck accidents occurring in the United States increases nearly every year. There were approximately 17 milion accidents involving motor vehicles in 1972 alone. A substantial number of the motor vechicles involved in accidents are automobiles and light trucks manufactured by respondent. In 1972, there were 86.4 millon automobiles registered in the United States; 41.1 milion or approximately 47.6% of these automobiles had been manufactured by GM. 9. Crash part comprise virtually the entire outer protective cover of an automobile or light truck and include the most frequently crash-damaged part. While any automobile or light truck part is susceptible to crash damage on occasion, crash parts collectively account for the preponderance of all automobile and light truck part replaced on account of crash damage. Unlike most other automobile and light truck parts, crash part are almost always replaced due to crash damage rather than due to maintenance or mechanical failure.

10. All servce GM crash part are and for many years have been produced either by GM or by independent manufacturers for GM. All of the relevant part are and for many years have been funnelled through GM for distribution. GM has and for some time has had and has intentionally maintained a monopoly and monopoly power over the distribution of these parts.

11. Unlike many other part it sells, GM for many years has sold and continues to sell service GM crash part exclusively to its franchise dealers who are located throughout the United States. GM' s franchise dealers, individually and in concert, have concurred , and urged upon GM, this policy of (4)sellng to them exclusively; and GM has acquiesced in and adopted this policy so as to extend to its franchise dealers, when wholesaling and installing the relevant 464 Complaint part, benefits of GM's monopoly position in the distribution of the relevant part. The dealers depend on and have for some time depended on GM as their sole source for new GM automobiles and light trucks and certain replacement parts applicable to GM-made vehicles. GM owns or has a substantial financial investment in a number of these dealers. GM franchise dealers either install the relevant part, wholesale them, or occasionally sell them to consumers. There are approximately 12 000 GM dealers in the United States, many of whom both wholesale and install the relevant parts. 12. GM franchise dealers wholesale and for many years have wholesaled service GM crash part principally to independent body shops (IBSs). There are approximately 30 000 IBSs in the United States. IBSs compete and have competed with GM dealers in installng the relevant part. Most of the relevant parts needed consumers are and for many years have been installed by GM dealers or by IBSs.

13. Because GM has distributed and sold the relevant parts exclusively to its dealers, IBSs have had to purchase said parts from the dealers and in so doing have frequently paid more for the parts than have competing GM dealers.

14. GM has refused to sell the relevant parts to its franchise dealers on equal terms. The dealers receive wholesale incentives on only those relevant parts which fit the lines of new cars the dealers are franchised to sell. This has effectively precluded many GM dealers from wholesaling additional relevant parts. 15. Service GM crash part are not installed in any vehicles other than those which have been assembled by GM. Furthermore, due to design proliferation by GM, any particular service GM crash part fits only one or at best a few models of GM vehicles. Thus, in excess of 000 different crash part were designed to fit GM automobiles and light trucks produced for sale in the U.S. during model years 1968- 1972.

16. Respondent, who has a monopoly in the distribution of service GM crash part, has engaged for some time, and is continuing to engage, in the following unfair methods of competition and unfair acts or practices, among other: (5) (a) refusing to sell the relevant part-goods which the IBSs are under a commerical compulsion to obtain--directly to IBSs or to any potential suppliers to IBSs other than GM franchise dealers; (b) bolstering its monopoly power through, among other things sellng the relevant parts exclusively to its franchise dealers; (c) adopting a method of distribution which substantially hinders Complaint 99 F.

competition in the distribution, wholesaling, and installng of the relevant parts;

(d) combining, agreeing or acting in concert with GM franchise dealers so as to substantially hinder competition in the distribution wholesaling, and installation of the relevant parts; (e) discouraging competition in the wholesaling of the relevant parts through utilization of disparate wholesaling incentives; (I) maintaining a method of distribution which provides GM with an unfair competitive advantage in the sale to its dealers of parts available from alternate sources; and (g) disseminating to GM franchise dealers lists which suggest the prices at which the relevant parts should be sold to installers and to members of the consuming public.

17. The effects of the acts, practices, methods, and power set forth in the preceding paragraph have been and are, among others (a) deter new entrants and raise barriers to entry into wholesaling and installng the relevant parts;

(b) enhance monopoly power and maintain monopoly pricing and inefficiency in the distribution of the relevant parts; (c) extend monopoly power and its effects in the distribution of the relevant parts to the wholesaling and installation of the relevant parts; (6) (d) curb efficiencies in the wholesaling of the relevant parts; (e) lessen competition in wholesaling the relevant parts; (I) restrain competition between GM dealers and IBSs in installing the relevant parts;

(g) restrain competition among GM dealers in wholesaling the relevant parts;

(h) increase prices to and otherwise disadvantage JESs in competing with dealer-owned body shops;

(i) decrease the availabilty of the relevant parts; (j) decrease competition in the sale to GM dealers of alternatesourced parts; and (k) increase prices to and otherwise disadvantage the consuming public.

18. The acts, practices and methods of competition alleged in this complaint, coupled with the monopoly power alleged herein, constitute unfair methods of competition and unfair acts or practices by respondents in violation of Section 5 of the amended Federal Trade Commission Act.

464 Initial Decision INITIAL DECISION By JOSEPH P. DUFRESNE, ADMINISTRATIVE LAW JUDGE SEPTEMBER 24, 1979 THE COMPLAINT 1. The Complaint is dated March 22, 1976, and charges that General Motors Corporation (GM) engaged in unfair methods of competition and unfair acts or practices in violation of Section 5 of the Federal Trade Commission Act (FTCA), as amended (15 U. 45), in connection with the distribution of "new service (2Jcrash parts" applicable to automobilies and light trucks assembled by GM (Complaint, Introductory Paragraph).

Crash parts are defined therein as:

. anyone Dr all of the following products: fenders, griles, bumpers, hoods, deck lids, doors, quartr panels, rear end panels, rocker panels, lamp assemblies, wheel opening panels, fender and rear end caps, tail gates, radiator supports and shrouds and mouldings, including inner and outer panels and all components of these products as well as all part necessary to attach the aforesaid to the boies of automobilies and light trucks. (Complaint l(d)).

The definitions of automobiles and light trucks are those generally understood, but are limited, respectively, to autos having seating capacity for no more than 10 persons and trucks having a gross vehicular weight of less than 10 000 pounds (Complaint l(a) and (b)).

2. Paragraphs 11 and 12 of the complaint reflect: that GM sells crash parts exclusively to its approximately 12 000 GM franchise dealers located throughout the United States and that the dealers either (1) install the parts themselves, (2) wholesale them, principally to the approximately 30 000 independent body shop operators (IBSs) in the U.S. who compete with the dealers in installng the parts, or (3) occasionally retail the parts to consumers. 3. One allegation is that IBSs must purchase the parts from their competitors, GM dealers, frequently at prices higher than those paid by the dealers (Complaint 13). Another is that since GM pays dealers wholesale compensation (explained below) only for crash parts for the brand of GM car the dealer sells (e. no wholesale compensation is paid to a Pontiac dealer who sells crash parts for a Buick), many GM dealers are effectively precluded from wholesaling crash parts for brands of GM autos and light trucks for which the dealer is not franchised (Complaint 14). 4. "Wholesale compensation" is a percentage of the list price GM Initial Deision 99 F.

pays to (or credits to) the dealer on his sales of crash parts to a businessman/repairer of damaged vehicles (i. an IBS or another dealer or commercial type purchaser but not to individual members of the public). In other words, wholesale compensation in the context of this case is a payment/rebate by GM to a dealer for performing a wholesaling function to the (3Jautomotive repair trade (Tr. 2005; CX 70IOB).' 5. Wholesale compensation is available to a GM dealer only when he sells parts applicable to the vehicles for which he is franchised (RA 795-799) to purchasers such as an IBS or, with certain limitations, another GM dealer. Wholesale compensation is not paid on sales to an indepedent wholesaler (CX 7813A-B; Tr. I0266) (CCPF 47).

6. The wholesale compensation allowance was and is designed to afford car dealers a satisfactory margin of profit (4Jon sales to IBSs and to encourage them to make crash parts available to the IBSs at the dealer price. On those part for which a wholesale compensation allowance is provided, the suggested general trade price is identical to the price the franchised dealer is to pay to GM. If a dealer adheres to the intent of the program, an IBS pays the dealer the same price as the dealer is charged by GM for crash parts used in the dealer body repair shop. (CX 70IOB).

7. Starting with an allegation that GM has a monopoly in the distribution of GM crash part Paragaph 16 charges that GM 1 The fol!owig abbrevitioIU will be us in thi iBion: - Trllript followed by the page number.

- Commision s Exhbit, followed by itb number. - Repondenw' Exbit, followed by its number. RA=d CCA - Repondents' and Commision counsls' AdmiBiona. ALX - Administrative Law Judg's Exhbit, fo\lowed by its number (Note: Thi device Wa/ us to inure that all paes of II document offered by Commiaion counae!; or that an exhibit the AL believed should be identified or included, bee a pa of the evidentiary rerd. For eJlample, ex 700A-G is typical. That eJthihit conswtu of Beven pael of II twenty-five (25) pae letter date May 12, 1967, from GM to Commision ota. Commision counsel declined to older the complete letter. In that intace, ALX-7 was ua to identify and place the rema.iningeighten (18) paes into evidence.

CCPF, CCB =d CC - Commision oounsel's Prpo Finding, Brief !ind Reply Brief. RPF, RB and RRB - Repondents' Propo Finding, Brief and Reply Brief. - Intervenor NADA's El'hibit, followed by its number. INPF, INB and INRB - Intervenor NADA' Propo Fidi, Brief and Reply Brief. IAF, lab and IARB - Intervenor AS' s Propo Findingt, Briefs and Reply Brief. 464 initial uelBlOn engaged/engages in the following unfair methods of competition and unfair acts or practices, among others, by: (8) refusing to selJ the relevant partgoos which the IBSs are under a commercial compulsion to obtain-directly to IBSs or to any potential suppliers to IBSs other than GM franchise dealers;

(b) bolstering its monopoly power through, among other things, selling the relevant part exclusively to its franchise dealers (abandoned or dismissed later-see below);

(c) adopting a method of distribution which substantially hinders competition in the distribution, wholesaling, and installing of the relevant part; Cd) combining, ageeing or acting in concert with GM franchise dealers so as to substantially hinder competition in the distribution, wholesaling, and installation of the relevant part;

(e) discouragng competition in the wholesaling of the relevant part through uti17..tion of disparate wholesaling incentives; CO maintaining a method of distribution which provides GM with an unfair competitive advantage in the sale to its dealers of part available from alternate sources (abandoned or dismised later-see below) and; (g) disseminating to GM franchise dealers lists which suggest the prices at which the relevant part should be sold to installers and to members of the consuming public (abandoned or dismised later-see below).

In Paragraph 17 it is alleged that:

The effects of the acts, practices, methods, and power set forth in the preceding paragaph have ben and are, among others, to (5) (8) deter new entrants and raise barriers to entry into wholesaling and installng the relevant part;

(b) enhance monopoly power and maintain monopoly pricing and ineffciency in the distribution of the relevant part;

(c) extend monopoly power and its effects in the distribution of the relevant part to the wholesaling and instaUation of the relevant part; (d) curb effciencies in the wholesaling of the relevant part; (e) lessen competition in wholesaling the relevant part; (f) restrain competition between GM dealers and IBSs in installng the relevant part;

(g) restrain competition among GM dealers in wholesling the relevant part; (h) increase prices to, and otherw disdvantae IBSs, in competing with dealerowned boy shops;

(i) decrease the availability of the relevant part; (j) decrease competition in the sale to GM dealers of alternatesourced part (abandoned or dismissed later-se below) and (k) increase prices to and otherw disadvantae the consuming public. The following order to Cease and Desist was proposed: Requiring GM to sell crash part, through and from whatever facilties it maintains to service its franchise dealers, to all vehicle dealers, independent boy shops and independent wholesalers at the same prices, terms and conditions of sale, sad prices to be subject to reasonable cost justified quantity discounts and stoking allowances. (Complaint, Notice of Contemplate Relief). Initial Decision 99 F. 10. However, in his Reply brief, Commission counsel proposed a different order. That version includes definitions of automobiles light trucks, crash parts, components of (6Jcrash parts service crash parts, service GM crash parts, independent wholesalers, independent quantity discounts and non-exclu- body shops, functional discounts, sionary terms or conditions of sale. Thereafter, Part I of the order calls for an end to GM's use of functional discounts, as defined. Part II calls for GM to sell its crash parts to all vehicle dealers independent body shops and independent wholesalers, as defined, at identical prices and on non-discriminatory, non-exclusionary terms except that ". . . graduated, non-cumulative, cost-justified volume and/or quantity discounts based solely on the sale of service crash parts" may be offered. Part III calls for submittal of a detailed plan to carry out the order no later than 90 days after the order is served on GM. Part IV calls for: (1) the plan to be put into effect 90 days after the Commission approves it; (2) notice to all GM customers for crash parts 30 days before a change takes effect and; (3) a notice in Automotive News or similar publication of each such change. Part V calls for an annual report to the Commission for five years on the anniversary of the date this Order becomes final "describing the manner of GM's compliance with parts I and II." (CCRB 132-136). 11. Commission counsel also moved in his Reply Brief (p. 98) for the acceptance of CX 7013A- , which is a letter dated March 5, 1976 from GM to the Director of the Commission s Bureau of Competition. However, the record for the reception of evidence is closed and I am not convinced that it need be reopened to receive the letter because the subjects in the letter either are addressed elsewhere or would not add critical or important evidence. (Note: Commission Rule 3.51(d) permits reopening the record by the ALJ to receive additional evidence but it is not appropriate in this instance. S ANSWER 12. GM answered the complaint on June 21, 1976, denying that the Commission had reason to believe that GM had engaged in unfair methods of competition and unfair acts or practices in distributing "new service crash parts, in violation of Section 5 of the Federal Trade Commission Act (FTCA), and denying that the proceeding would be in the public interest. To the numbered paragraphs of the complaint, GM:

(1) denied the accuracy and applicability to the proceeding of all definitions in the complaint except the one for automobiles (Answer U);

(g) 464 Initial Decision (2) admitted manufacturing and sellng a wide variety of products, including automobiles and trucks, (Answer, n 3); (3) admitted sales in calendar year 1972 of $30.4 billion with after-tax profits of $2.16 bilion, (7)and total assets of $18.3 bilion compared with 1975 sales of $35.7 bilion with after tax profits of $1.3 bilion (Answer, n 4);

(4) admitted being the largest manufacturer of automobiles and light trucks in the U. , (Answer, n 1 4 and 5); (5) admitted that in 1972 GM sales of automobile replacement parts including "crash" parts exceeded $250 milion (Answer, n 6); (6) admitted that GM engages in commerce and affects commerce (Answer n 7);

(7) admitted that of the 86.4 milion automobiles in U.S. operation in 1972, approximately 41.1 milion had been manufactured by GM or its subsidiaries (Answer, n 8);

(8) admitted "(a)ll service GM crash parts are, and for many years have been produced either by GM or by independent manufacturers for GM, . . . and have been funnelled through GM for distribution " but denied that it had or has ". . . intentionally maintained a monopoly and monopoly power over the distribution of these parts." (Answer, n 10);

(9) admitted that it sells new GM crash parts exclusively to the approximately 12 000 GM dealers, some of which it owns or in which it has a financial investment, who either install or sell the parts, or do both (Answer, n 11);

(10) admitted that any particular new GM crash part may not fit all models of GM vehicles and that in excess of 5 000 different service GM parts fi GM autos and trucks for model years 1968-1972 (Answer, n 15).

GM either denied the remaining allegations or stated that it was without knowledge or information suffcient to form a belief regarding their truth.

13. As noted above, on pages 4-5, Commission counsel later abandoned or agreed to the dismissal of paragraphs 16(b), (I) and and 17(j). (See "Order (1) Dismissing Paragraphs 16(b) and 17(j) Of The Complaint, and (2) Denying GM Motion For Interim Rulings To Guide Further Hearings" dated September 29 , 1978; Tr. 10581). (8) THE INTERVENORS 14. The National Automobile Dealers Association (NADA), which had on July 7, 1976, 8 690 members who were GM dealers, was permitted to intervene by order dated January 11, 1977. Counsel to Initial Decision 99 F. NADA has participated in the trial by questioning witnesses, callng his own witnesses, offering exhibits, making oral arguments, and submitting proposed findings and briefs.

15. In addition, the Automobile Service Councils, Inc. (ASC) which had over 2 000 independent body shop operators as members on October 3, 1978, was permitted to intervene by order dated October 16, 1978. Counsel to ASC has participated by filing briefs. THE HEARINGS 16. Prehearing conferences were held in Washington, D.C. on April 7, July 29, September 22, October 29, and November 23, 1977 and on January 24 1978 (CCPF, p. 1).

17. Both parties and intervenor NADA exchanged trial briefs in support of their respective positions. These included legal arguments and lists with copies of proposed exhibits and the names of witnesses with short narrative summaries of expected testimony. 18. The hearings began in Washington, D. , on May 15, 1978. The record for the reception of evidence was closed on May 22, 1979. In all, there were 82 days on which hearings were held, including an inspection of the GM parts warehouse in Baltimore, Md. There are approximately 16 285 pages of transcript. (Note: There are some gaps in pagination due to changes from CCroutine" to "daily" or from daily" to " rush" copy, Feb. 2- , 1979, April 18, 1979. When such changes occur the reporter must estimate the number of pages required for transcription of notes. If the estimate is Iowa gap in pagination results.

19. Sixty witnesses testified for the Commission, 21 testified for , and 3 testified for NADA (RB 1). Of these, 24 were IBS witnesses from the following seven trade areas: Buffalo, New York; Mansfield Ohio; Cleveland, Ohio; New Orleans, Louisiana; St. Louis, Missiouri; Spokane, Washington; and Tucson, Arizona (ALJX 26). In addition testimony of two IBS witnesses was stipulated (CCPF 106; RRB 106). THE STRIKING OF TESTIMONY AND REJECTION OF EXHIBITS 20. The testimony of four GM witnesses was stricken because counsel for GM declined to observe my order that they were to hand over to Commission counsel pretrial reports of intervews (9Jof witnesses. My "Order Granting Motion of General Motors Corporation for Production of Interview Reports" dated April 10, 1978, called for each side to provide the other with ". . . all interview reports relied upon in connection with the witness s testimony" one week before a witness testified. The purposes of the order were to have 464 Initial Decision each side apprise the other, within reason, of testimony to be elicited by Commission counsel in connection with the allegations in the complaint, the defenses of GM, to encourage counsel to execute stipulations, and otherwise to expedite the trial (Tr. 10619-20). The order was discussed at considerable length at the hearing Septemter 27, 1978 (Tr. 10581-10629; 10672-10693). Commission counsel made it clear that he had maximally complied with the order (Tr. 10624-25).

21. On that date, in the course of the hearing (Tr. 10582) and later in connection with the testimony of GM witnesses Cann on October 2 (Tr. 11119-11179), Mack on October 5 (Tr. 11555-11701), Faulkner on October 6 (Tr. 11718-11817), and Vulbrock on October 17 (Tr. 12280-12354), Commission counsel raised the question as to whether any intervew reports existed.

22. Counsel for GM said that he had no document/interview reports within the purvew of the order and that, even if he did, an AW lacked authority to compel what the April 10 order required. Thereafter, Counsel for GM showed me some papers/notes. After examining them I concluded they were intervew reports within the scope of the order. However, Counsel for GM continued to decline to hand over copies to Commission counsel. After I made some handwritten marks on them to identify those portions deemed privileged, the report were returned to counsel for GM. The "Order Denying Motion for Reconsideration of Order of September 27 1978 Requiring the Production of Interview Reports" dated October 13 1978, elaborates on the action taken at the hearing. 23. The "Order Modifying Order Granting Motion of General Motors Corporation for Production of Interview Reports" dated October 31, 1978, ordered the striking of the testimony of the four GM witnesses. (See Commission Rule 3.38). Thereafter, the four documents were placed in a sealed envelope and delivered to the Commission s Secretary so that they may be examined by the Commission upon its review. (See "Order Re In Camera Documents Delivered to the Commission s Secretary" dated Feb. 14, 1979). 24. The testimony (Tr. 875(4787) of a Commission witness whose name and testimony are in camera at his and Commission counsel's request was stricken as being cumulative (Tr. 8787) (CCRB TI 2). Part of the testimony of Commission witness Perschall were stricken for a time because the parts were unreliable, due to their being based on documents prepared by another person, with these foundation documents either not produced at the hearing or not being credible. However, the (lOjtestimony was reinstated without objection by counsel for GM after underlying documents were Initial Decision 99 F. provided. (See "Order Reinstating Stricken Testimony of Commission Witness Kenneth Perschall" dated May 25, 1979). 25. In accord with Commission Rule 3.43(g), the rejected exhibits and testimony remain a part of the offcial record, although they have not been considered in reaching or preparing this Initial Decision.

BASES FOR THE FINDINGS OF FACT 26. The following findings of fact are based on a review of the allegations made in the complaint, respondent's answers, the documentary evidence, and consideration of the demeanor of the witnesses.

27. The proposed findings of fact, conclusions, and proposed orders, together with reasons and briefs in support thereof fied by each side and by the intervenors have been given careful consideration. Many proposed findings have been adopted as submitted or in substance. To the extent not adopted by this decision in the form proposed or in substance, they are rejected. Further, any motions not ruled upon are denied.

28. For convenience, the findings of fact include references to supporting evidentiary items in the record. Such references are intended to serve as guides to the testimony, evidence, and exhibits supporting the findings of fact. They do not necessarily represent complete summaries of the evidence considered in arriving at such findings.

FINDINGS OF FACT Background 29. The Commission investigation of the distribution and sale of etc.crash parts which, essentially, are fenders, gIiles, moldings, began in the mid-1960' s. Operators of IBSs had complained to the Commission that automobile dealers were charging them excessive prices for crash parts, thereby making it diffcult to compete with dealers for collision repair business. The IBSs sought to buy crash parts at the same price that automobile dealers paid for parts used in their own repair shops (ALJX 14M, Supp. to CX 7014). 30. Prior to September 12, 1967, GM simply provided its dealers with a suggested general trade price to be charged wholesale purchasers of crash parts (CX 7015A). GM estimates that under that system wholesaling dealers allowed purchasers at wholesale an average discount in excess of 18% from the list price (CX 70l5B). 31. On September 12, 1967, GM proposed a plan to its (llJdealers g., 464 Initial Decision callng for the payment of an overriding discount of 12% from dealer price to any GM new car dealer on a "qualified wholesale sale" of seventeen (17) categories of crash parts, including: fenders; grilles; bumpers; radiator supports; and body side moldings. A "qualified wholesale sale" would be the sale and delivery (less returns and repurchases by the dealer) of such parts to: Automotive repair shops, automotive body shops and gasoline servce stations which purchase the eligible General Motors part for the repair, rebuilding or servicing of General Motors vehicles for such purchaser s retail and service customers, except any such purchaser in which the sellng dealer, or any stockholder or principal thereof owns or controls any financial interest (CX 7015C). 32. In February, 1968, the Commission advised GM that it intended to bring suit to bring about price parity between franchised car dealers and independent body shops (AUX 14N, Supp. to CX 7014z) 33. After negotiating with Commission staff and prior to adopting the wholesale compensation plan, GM stated: General Motors' cost would he increased by the amount of the discount, by the cost of administering the program to insure against fraudulent claims (by GM dealers), and by the costs of the routine paperwork to administer the program. All of those additional costs would have to be factored into prices for these part resulting in significant price increases to the consumer (Emphasis added) (eX 700E; eCPF 299). In other words, it was the judgment of GM's top management that if wholesale compensation were necessry to save the basic structure of GM's distribution system the overall advantages outweighed the costs of wholesale compensation (ALJX 7 and 8)(RRB 299).

34. After many discussions with Commission staff, GM and the three other principal U.S. auto makers agreed in the fall of 1968 to implement a wholesale compensation plan (CX 701OD). As early November 21, 1966, Commission staff had told GM representatives that the basic reason for the investigation was ". . . to require General Motors to distribute its captive sheet metal parts (i. , crash parts) on the same terms as it presently distributes its competitive parts sparkplugs, fiters, etc. (RX 26A). 35. The Commission announced on October 22, 1968, that "the leading automobile manufacturers" had agreed to adopt such a (12) wholesale compensation plan for crash parts . . . to help overcome what the Commission considers to have been competitive disadvantages facing independent auto body repair shops." GM put the plan into effect with the introduction of the 1969 models (AUX 14N Supp. to CX 7014).

36. The Plan did not provide for an allowance when a dealer Initial Decision 99 F. wholesaled a crash part intended for a light-duty truck or for several passenger car lines, including the Chevrolet Corvette, Vega and Monza, Pontiac Astre, Oldsmobile Starfire, and Buick Opel and Skyhawk. This was because no imported car manufacturer or distributor was known to make a wholesale compensation allowance available to its dealers and those lines/makes were considered to compete for the most part with imports (CX 701OC). Approximately 15 different foreign car manufacturers sell new cars in the United States (CCA 43).

37. GM pointed out that if GM made such an allowance available and the manufacturers of foreign makes did not, one of two things could happen: (1) It could make no change in the dealer net price, in which case GM would have to absorb the cost of a 25 percent allowance on the wholesale portion of its crash parts business in the excepted car lines Corvette, Vega, Monza etc. which on the basis of 1974 sales at 1974 prices, would cost at least $17 millon per year; or (2) GM could increase the price for crash parts, in which case they would cost more than a similar part for a competing import (CS 701OD).

38. Under the plan adopted, average wholesale compensation was 23% plus a possible additional 5% Stock Order (PAD) Allowance (described at p. 37) applied to purchases (CS 7018). On October 1 1968, GM' s prices on the specified parts were increased to enable GM to recover the amount that would be lost from making the increased wholesale compensation payments (CX 7022A). 39. In a January 31, 1969, GM review (CX 7021A and B) the following were noted:

(1) The Service Section, in collaboration with representatives from each of the five car divisions, established overriding discounts on the categories of part selected by the Federal Trade Commission;

(2) The discounts ranged from a minimum of 20% on high value top and quarter panels to a maximum of 25% on nine categories of part priced under $20.00 at the dealer level. Overriding discounts of 22% and 23% were applied to the remaining categories such as fenders, hoods, and deck lids. The average overriding discount for all of the part was estimated to be 23%; however due to some (13) changes in price levels, the actual rate for the five car divisions was 22.4%;

(3) The financial effect on General Motors of adding wholesale compensation on 10 208 parts in the FTC selected categories was estimated to be slightly over $18 000 000, computed on 1966 volume. 464 Initial Decision This amount was recovered by increasing dealer prices 1.40/0 applicable to all parts including crash type items; (4) Dealers' gross profit dollars had been reduced 18.4% but the percent of profit increased from 26.9% to 27.4% (CX 7021A); (5) The amount of decrease in gross profit dollars and percent of profit varied among divisions as illustrated in the following table: Gross Profit Amount Gross Profit Percent Millons Change Change Old New New vs. Old Old New New vs. Old Chevrolet 134. 114. (19. 89) (14. 25.9 27. 1.4 Pontiac 157. 118.45 (38.68) (24. 28. 27.4 (1. Oldsmobile 147. 112. (35.58) (24. 28.5 27. Buick 190. 146. (43.65) (22. 27.9 27. Cadillac 192. 178. (13.34) ( 7. 24.5 27. Total 822. 671. (151. 14) (18.4) 26. 7.4 Both Chevrolet and Cadilac were affected to a lesser degree than the other three divisions because of a difference in the pricing and discount patterns of the old program, whereas under the new program, a uniform pattern applicable to all five car divisions was adopted.

(6) Dealers had been operating in the area of a 25% gross profit on wholesale sales for the previous three years and dealers would show an increase in profit dollars on their retail sales due to increases in parts prices;

(7) Since dealer expenses had been increasing and there was a need to encourage dealers to engage in the wholesaling of crash parts, an increase in the wholesale compensation allowance might be in order. The cost to General Motors was estimated (14)to be approximately $850 000 for each additional percent allowed (CX 7021B).

40. In the fall of 1969, GM's suggested list prices on both replacement and crash parts again were increased. The increases averaged approximately 4%, with no part raised in excess of 6 Y,%. Changes also were made in base discounts to dealers from GM's list prices. It was forecast that the change in the base discount rate would result in a slight reduction in the dealers' average gross margin on retail sales, but that this would be more than offset by an increase in wholesale compensation. Dealers began to receive a base discount of 40% and a standard wholesale compensation rate of 25%. Previously the discounts had ranged from 35% to 44% and wholesale compensation rates had varied from 20% to 23% , depending upon Initial Decision 99 F. the parts groups involved. All part numbers within each compensable parts group were made eligible for wholesale compensation (CX 7023A-B).

41. Generally, each increase in the dealer price due to changes in the rate of wholesale compensation was accompanied by a proportionate increase in the list price of the relevant parts (RA 904, RA 905). This was due to GM's maintaining the dealer price at approximately 40% ofthe list price (CX 7225C) (CCPF 304). 42. Crash parts were made eligible for both the 5% Stock Order (PAD) Discount and wholesale parts compensation. The object was to assist dealers in greater penetration of the wholesale parts market both replacement and crash, which was estimated to be approximately six times larger than the total retail parts market. As an example a carburetor, Group 3.725, having a compensation rate of 12. (which indicated that only approximately 50% of the volume of parts in that group were formerly eligible for wholesale compensation) was made eligible for 25% wholesale compensation in that group (CX 7023A-B).

43. The "General Motors Parts Division Body Shop Price Schedule" (CX 7422A-Z-3) contains a suggested list price and a suggested trade price for the automotive replacement parts GM offers for sale to its dealers (Tr. 2056). The list includes all of the parts which are in the definition of crash parts contained in the complaint in this matter (Tr. 2056). The dealer net price, before other discounts or rebates, is the same as the suggested trade price and is the price GM recommends dealers charge to IBSs and other commercial auto body repairers (Tr. 2059).

44. In 1970, another FTC investigation of the effects of wholesale compensation began (RX 28F). The staff concluded: (1) that wholesale compensation had not achieved price parity between dealers and IBSs; (2) that prices to consumers had risen; and (3) that there would be an estimated 10% drop in consumer prices if wholesale compensation ended (RX 28G). (15) 45. On March 21, 1972, still another investigation was initiated to look into " . . . possible monopolization of crash parts by the auto makers" (RX 28H). There were many discussions within GM regarding ways to resolve the controversy with the FTC (Tr. 2224). Two objectives were to settle the controversy and to reduce or eliminate wholesale compensation costs (Tr. 2143, 2168) (CX 7253) (RPF 259).

46. On July 11 , 1975, a settlement proposal was made to Commission staff under which GM would sell crash parts directly to any IBS at the dealer net price from the 27 field warehouses 464 Initial Decision operated throughout the United States if that proposal would end the investigation. GM acknowledged that the Robinson-Patman Act (15 U. C. 13) would require that the independent bddy shop operators be accorded the same prices and services as the dealers (CX 701OE).

47. The proposal called for car dealers to continue to distribute the overwhelming majority of crash parts to independent auto body repair shops and to receive wholesale compensation from GM for substitute its 27 warehousedoing so. GM did not intend to distribution locations for the dealers' 12 000. Rather, the warehouses would be an alternative source when an IBS operator felt that he could not buy the crash part he needed at a competitive price from a dealer (ALJX 13D, Supp. to CX 7012).

48. Also, in the summer of 1975, the Commission retained toCambridge Management Associates, management consultants, survey existing warehouse distributors to determine whether they would be interested in selling crash parts. The survey contemplated that the distributors would sell to jobbers, who in turn would sell to the IBSs or to car dealers who chose to buy from them rather than jobber-andirectly from the manufacturer. It was projected that the additional link in the distributive chain-would need an average gross margin of 25%-46% to perform his function (ALJX 14S, Supp. to CX 7014).

49. In October, 1975, the Automotive Warehouse Distributors bulkAssociation (A WDA) whose membership accounts for the great of independent wholesale parts distributors, advised the Commission that its members were not interested in distribution of sheet metal parts (i. crash parts) and that they were not equipped to stock and handle them (ALJX 14T, Supp. to CX 7014). AWDA is a trade association representing approximately 500 of the nation s 1000- 1500 warehouse distributors (WDs). GM is also a member of AWDA. WDs sell automotive replacement parts to jobbers who resell them to installers (IX 2; Tr. 2248; 8640-1) (CCPF 322). 50. AWDA opposed IW-IBS access in order to prevent its WD members from facing additional competition (Tr. 12588). The (16) members feared that GM would sell maintenance-type (replacement) parts, as well as crash parts, to independents and that this would place GM in direct competition with WDs for jobber business. If GM sold crash parts only to IBSs, AWDA would not oppose IW-IBS access (Tr. 12591-92; 12664-6). (CCPF 323). A former executive of A WDA testified that changing the distribution system might reduce the availability of crash parts (Tr. 12523-24) (RRB 323). 51. In early October, 1975, GM raised the amount of the g., Initial Decision 99 F. wholesale compensation allowance from 25% to 30% (ALJX 13G Supp. to CX 7012).

52. Later, on February 5, 1976, in another proposal to the Commission, GM said that it would broaden the Wholesale Compensation Plan by paying a dealer an allowance for the sale of eligible crash parts to an IBS for any make of GM car a Pontiac dealer than would be able to obtain wholesale compensation on the sale of Chevrolet, Oldsmobile, Buick and Cadilac crash parts (ALJX 13G Supp. to CX 7012). Under this proposal, any of the 12 000 GM dealers would have been able to claim wholesale compensation whenever an eligible crash part, regardless of the make of car it fit, was sold to an IBS (ALJX 13H, Supp. to CX 7012). This proposal was made after GM offcials concluded that the July 11, proposal would not be acceptable (ALJX 13B, Supp. to CX 7012).

53. In addition, the proposal called for GM to make crash parts for subcompacts and light duty trucks eligible for wholesale compensation on a trial basis, with the option of changing at the end of six months if GM's principal competitors, including the foreign manufacturers, did not implement a similar plan (ALJX 131, Supp. to CX 7012).

54. Also in 1976, on March 1, 8, and 12, the Subcommittee For Consumers of the Committee on Commerce of the United States Senate conducted hearings on the cost of automobile crash parts and subsequently published "Automobile Crash Parts, the transcript of the hearings for the use of the Committee on Commerce (ALJX 17). At the hearings the Director of the Commission s Bureau of Competition testified that GM's February 1976, proposal was "particularly disappointing" (RX 28,), "totally inadequate" and that the wholesale compensation plan had "raised prices to consumers without achieving its goal of price parity for IBSs" (RX 28K). He also commented that State Farm Mutual Automobile Insurance Company favored requiring the auto makers to sell directly to independent wholesalers (RX 28I--).

55. Later that month the February proposal by GM was rejected by the Commission and the Complaint was issued. See p. 1. (17) The Respondent 56. General Motors Corporation is a Delaware corporation organized on October 13, 19I6, with its headquarters at 3044 W. Grand Boulevard, Detroit, Michigan. (Answer, n 2). GM is the successor to the General Motors Company, which was organized on September , 1908 (RX 31OR).

464 Initial Decision 57. GM is, and for many years has been . engaged in the manufacturer and sale of a wide variety of products, including automobiles, trucks, buses, diesel locomotives, diesel engines, earth moving equipment, household appliances, and automotive parts (Answer, U 3). The principal makes of automobiles GM manufacturers and sells in the United States are: Chevrolets, Pontiacs Oldsmobiles, Buicks, and Cadilacs (RA 746 and 748). There is a separate GM division for each of these, with each division franchising dealers to sell its own make of auto (Tr. 1999). Trucks are franchised by the Chevrolet Division and the General Motors Truck and Coach Division (Tr. 1999). A dealer franchised to sell several makes is franchised by each division whose brand he sells (Tr. 2000- 2001).

58. GM and its dealers' primary business and interest is in sellng cars and trucks (Tr. 9869; 12651) (CCPF 183). 59. New car customers expect the manufacturer of the car that they buy to see to it that parts and service are available for that car (Tr. 11008). As a consequence, car manufacturers such as GM must provide the necessary back up stocks of parts (Tr. 11009; 13978-79). Many other considerations such as the car s styling, size, sticker price, gas mileage, and resale value are at least as important to a prospective car purchaser as the cost of crash parts (Tr. 1382) (CCPF 185).

60. Accident repair costs arc of little, if any, importance to most purchasers buying an automobile since they believe their insurance will cover damage expenses above the deductible amount (CX 7815P) (CCPF 188).

61. GM automobiles come in 12 different body sizes. As many as four car divisions produce unique models of the same body size. Generally, each body size and each division s model of that body size consists of both unique and common crash part Chevrolet' s Bbody Impala and Caprice (Tr. 10124-30, 10153-55) (CCPF 13) (RRB 13).

62. In 1976, GM-manufactured automobiles accounted for 45. of total U.S. automobile registrations and GM trucks for 42% of total S. truck registrations (CX 7409A, C).

63. GM also sells various maintenance type automotive replacement parts such as wire and cables, spark plugs, brake shoes batteries, and carburetors to nondealer resellers who (18Jsell installers, including some GM dealers (RA 757, 758, 761). 64. The following table shows in gr- )f detail and for more recent years the level of operations of GM (RX 310P and Q): 694 (1968) 252,468276,076 928276349355 002,451 629,617663191 728198 399113 315581(1968) 005829(1968) 676 1967 026627 084 529 912 1 1 5,634 $20 $ $ $ 231,414806765 270066301 812194 037584736223 759,543 284825 778124 962,316 0500851972 12,928 162 273 876 940668 2, 1 8 $30,436 $ $ $ 272,516527231 928261714370 884600 930827777726 848796 384245 258583 436220 3730101977 9613,337 9441 1,379 8701270 $54 $ $ $ $$15 U. Special and for .the In Manufactured Plant Employees Manufactured Stock: for(Excluding of U. Retained on Stockholders Coaches Trucks the U. Business: Worldwide No. of & in Manufactured the and Sales Income Income PreferredCommon Use EquipmentTools) Worldwide in outside Net Net Dividends Net Expenditures Payrolls Average Number Cars Trucks Cars 464 Initial Decision 65. GM is the largest manufacturer of automobiles and light trucks in the United States (Answer, n 5). The only light trucks which GM manufactures and sells in the United States are made under the Chevrolet and GMC names (RA 750). 66. In 1972, GM sales of automotive replacement parts, including crash parts as defined in the complaint, exceeded 250 millon dollars (Answer, n 6).

67. GM engages in a continuous and substantial course of trade in commerce and affects commerce throughout the United States (Answer, n 7).

Crash Parts 68. All service GM crash parts, as defined in the complaint, are and for many years have been, produced either by GM or by independent manufacturers for GM. They are distributed by GM exclusively to its dealers who either wholesale, otherwise resell, or install the part (Answer, n 10). Michael C. Mehan, Executive in charge of GM's U.S. servce parts operations General Motors Part Division (GMPD) and AC-Delco Division, testified that the parts enumerated in the complaint are crash part but did not agree that the definition was all inclusive (Tr. 2009). 69. GM sells most of its crash parts exclusively to approximately 000 GM dealers (Answer, n 11). However, not all GM dealers have body shops (CCA 61).

70. As of December 31, 1974, GM owned and operated 23 GM dealerships and had a temporary financial interest in 379 (RA 764). Some of the 23 and many of the 379 conducted body shop operations (RA 767 and 769) and many of each category resold crash part to installers in 1974 (RA 768A, 770A).

71. There are more than 5 000 different crash parts for GM 1968- 72 model year automobiles and trucks (Answer, n 15). There are 112 different Chevrolet fenders. In recent years Chevrolet has sold an average of 4 500 of each different fender in each year. Since there were approximately 6 500 Chevrolet dealers (as of May 12, 1967), Chevrolet dealers buy an average of less than one fender per year of each different fender (CX 7000D).

72. All crash part produced by independent manufacturers for GM are produced from tooling GM owns except for parts for (19)step vans (RA 720).

73. On eligible parts dealers currently receive a rebate of 30% (wholesale compensation) of the dealer price on qualified sales to wholesale customers, including les (Tr. 2068-70, 10252-54, 10294). Initial Deision 99 F.

74. Crash parts as a class are bulky and require considerably more space for storage than replacement parts. They are more easily damaged and harder to handle than the typical replacement part (Tr. 12509-10; 12629-30; 10459-61; RX 51S) (RPF 79). 75. GM does not pay wholesale compensation on dealer resales of crash parts exclusively used on a brand of automobile the dealer is not franchised to sell a Chevrolet dealer sellng a crash part for a Pontiac (RA 795-799).

76. GM pays wholesale compensation on resales of eligible crash parts by its dealers: (1) to independent automotive repair shops, body shops, gasoline stations, fleet users (5 car or light truck minimum); (2) to non-GM-dealer new car or truck dealers Ford, American Motors, British Leyland; (3) on emergency sales to another GM dealer to meet a service customer s needs or to an outlying GM dealer who has been approved as a buyer by the cognizant GM franchising division of GMPD; (4) to independent used car and truck dealers; (5) federal, state, county, and municipal government agencies. Sales: (1) to retail customers, (2) to insurance companies for use on a vehicle owned or titled in the name of an insured, (3) to a department of the sellng dealer s dealership, or (4) to anyone who purchased eligible parts for resale directly or indirectly to a GM car or truck dealer, are typical of those resales for which wholesale compensation is not to be claimed (RA 801, Attachment A).

77. In order to get his rebate and after he has wholesaled the part, the dealer fies a report in which compensation is claimed (Tr. 2006). Wholesale compensation is not to be paid to a dealer for a part used in the dealer s own body shop (Tr. 4753; Tr. 11020) and is to be paid on dealer-to-dealer sales only if an emergency existed or if the sale had prior GM approval (e. to an isolated dealer in a remote location) (Tr. 2087-90; CX 7253D) (RPF 34). 78. Wholesale compensation payments require substantial administrative expenses in addition to the cost of the payments. For a GM dealer to obtain wholesale compensation, he must obtain a form from GM, keep track of qualified sales, enter the sales on the form and send the form to GM. GM must transmit the form to the dealer receive it back, process it, issue credits, and, on occasion, audit dealers (CX 700E; Tr. 2165-6 11972-74) (CCPF 307). (20) 79. The added costs of wholesale compensation are not offset by any other benefits, such as the receipt of superior servce. GM dealers do not perform any services under the wholesale compensation plan that they did not perform prior to its adoption (Tr. 10231). To receive wholesale compensation on crash parts, GM dealers need not stock, maintain any facilities, deliver, solicit sales, pass out 464 Initial Decision technical bulletins, or do anything else except sell to qualified purchasers (Tr. 2201, 10239--0) (CCPF 309). 80. Except for parts carried over for use as original equipment in successive model years, service GM crash parts typically are inventoried, for 7 to 12 years (RA 864).

81. GM does not rechrome or sell rechromed bumpers, or salvaged/used crash parts (RA 867-869).

82. In 1975, more than 13 000 crash parts were included in GM' wholesale compensation plan (RA 893).

83. When GM increases dealer prices, it usually also increases suggested list prices so that the discount from suggested list to dealer net price remains approximately the same as before the price change (RA 905).

84. GM usually starts production of a 6 to 8 months' supply of crash parts for automobiles and light trucks at least 2 months prior to introducing the new model (RA 913 and 916). General Motors Dealerships 85. The basic policy of GM is and has been to distribute its cars through independently owned and operated dealerships (CX 7029B). There are about 5 000 GM dealers who wholesale GM crash parts (Tr. 10285). There are approximately 7 000 additional who could purchase and wholesale crash parts (RX 2). (RPF 183). 86. As of February 24, 1977, in the United States the various divisions making GM cars had the number of dealers shown: Cheverolet-5992; Pontiac-3239; Oldsmobile-3322; Buick-3025 and; Cadillac-1616 (RX 33B-F).

87. GM at one time owned automobile dealerships in the Manhattan section of New York City but these were phased out in 1976. GM currently owns and operates no automobile dealerships but does own 18 truck dealerships, which are primarily involved in the sale and servce of medium and heavy duty trucks (Tr. 9863-4) (RPF 186).

88. To facilitate the opening of dealerships, GM has a Dealer Investment Plan operated by its Motors Holding Division (21)(MHD) (CX 7029C). As of October 31, 1964, of 13 395 dealerships, 306 or 3% of the total, were operating under the MHD Plan (CX 7029C). 89. As of December 31, 1977, GM had a financial interest in 345 or 3% of the 11 660 GM dealerships in business (RX 34). 90. MHD dealerships accounted for an estimated 4.3% of total GM dealer wholesale parts sales and an estimated 5.3% of total GM dealer body shop sales (RX 34; RX 38; RX 40) (RPF 187(d)). . . .

Initial Decision 99 F. 91. Under the plan the entrepreneur/dealer provides at least 25% of the capital required and MHD provides the balance. The dealer gets voting stock and is paid a salary. He buys MHD-owned stock in the dealership in order to increase his interest (CX 7029D- E). The dealer-operator conducts the day to day operations (Tr. 9846- 47; CX 7029E). However, two MHD representatives sit on the board of directors (CX 7029E). Between 1929 when the plan began and 1964, 1 139 dealers bought their businesses in this way (CX 7029F). 92. The interest of MHD is limited in duration. Between 1970 and 1977, the average length of time during which MHD held financial interests in such dealership ranged from five years, three months, to six years, seven months (RX 34). 93. MHD financed dealers are free to buy parts from any source they choose and otherwise to operate their dealerships as they see fit. Purchases of parts from GMPD are on the same terms and conditions as those made by other dealers and any resales are made at dealer-chosen prices. Such dealers are audited periodically to determine the status of MHD's investment and to advise the dealers concerning efficiency of their sales and other methods of operation (CX 7029G).

94. Each car manufacturer has the same incentives GM has to distribute crash parts in the most effcient way possible (Tr. 15751- 55; 15794-95) (RPF 102). All U.S. auto manufacturers and all foreign automobile companies that have been selling cars in the U.S. since the early 1960's now distribute crash parts basically in the same way (Tr. 2223; Tr. 9870) (RPF 103). But see "Chrysler s Mopar Experi. ence, page 36.

95. The terms of the contract between GM and its various dealers are set forth in a "Dealer Sales And Service Agreement". The agreement usually is executed for GM by the general sales manager of the division responsible for the make of car the dealer sells another GM offcial, and a partner or the proprietor of the dealership (RX 228-13). The contract calls for the dealer to " carry in stock at all times an inventory of Parts and Accessories adequate to meet customer demands and for warranty repairs special policy adjustments and campaign corrections. . ." Neither dollar amounts nor number of (22Jitems to be inventoried is specified in the contract (RX 2-W). The agreement also authorizes GM personnel to examine and audit the books of the dealership (RX 2-Z). 96. GM makes recommendations as to the space which should devoted to the parts department which are based on the dealer anticipated monthly sales of cars ("Planning Potential") and Net Dollar Inventory. The following table illustrates this: 464 Initial Decision Plannlng Potential Sq. Ft. Net Dollar Invento Sq. Ft. 1200 $ 10 000 1500 101.125 1700 000 4500 201-225 2200 000 8500 301-350 3300 150 000 14000 (CX 7234B) 97. GM also recommends a mix of parts inventory on a model year basis which its dealers should have. The optimum pattern per GM calls for the inventory to be broken down as follows: Year Current Model 10% 1 Year Old 20% 2 Years Old 25% 3 Years Old 20% 4 Years Old 15% 5 Years or Older 10% (CX 7274E) 98. The dealer is required to use and keep up to date a satisfactory uniform accounting system of a type designated by and to furnish to GM". . . by the tenth of each month, complete and accurate financial and operating statements. . ." (RX 24). GM audits some dealers claims under the wholesale compensation plan (RA 806) and reviews the records of all of its dealers (RA sila). 99. In wholesaling eligible service GM crash parts a GM dealer can obtain 30% lower prices on parts that are uniquely applicable to the make of automobile he sells than can GM franchise dealers who are franchised to sell other makes (see RA 923A). 100. GM dealers and IBSs perform most of the body repair work which is done on GM automobiles and light trucks (23j(Tr. 1202). GM dealers are the principal competitors of IBSs in performing such repairs (RA 774-; Tr. 1201-(2). IBSs do body work on all makes of vehicles but most GM dealers tend to specialize in repairing the models they sell (Tr. 1200-- 9495 9498) (CCPF 39). 101. The price GM dealers charge IBSs for new GM crash parts may vary throughout the United States (CCA 55). GM Dealers as Parts Wholesalers 102. GM dealers are not expected to stock all or even some of the slowest moving crash part (Tr. 10086-8). In states where there is an inventory tax, GM dealers do not want these slower-moving parts , 490 FEDERAL TRE COMMISSION DECISIONS Initial Deision 99 F. T. to sit on their shelves. They prefer to carry faster-moving items (Tr. 6982-3) (RPF 77).

103. GM dealers need not engage in parts wholesaling or the operation of a body shop (Tr. 9094). If they wholesale part, they are not required to stock part (Tr. 9904-5; 10268). The dealer determines which part if any he wil stock for any of his operations. GMPD does not control the inventory practices of GM dealers (Tr. 9094; 9907) (CCPF 36).

104. Dealers usually do not wholesale crash part for makes of cars they do not sell (Tr. 2126). If wholesale compensation were paid on such sales some dealers would enter that field and others would drop out (Tr. 2132).

105. Each of the crash part listed in paragraph 1D of the complaint is eligible for wholesale compensation (Tr. 2019). 106. A GM dealer is at an automatic 30% price disadvantage in wholesaling non-franchise crash part those for makes he is not franchised to sell (Tr. 10263-4, 10266). This constitutes a near total entry barrier to such sales (Tr. 14015), as relatively few GM dealers do in fact wholesale crash part for makes for which they are not franchised (Tr. 2126) (CCPF 255).

107. It is the position of NADA that elimination of this limitation would benefit IBSs and increase availabilty of parts to consumers at lower cost by increasing the competition among franchised dealers sellng different model vehicles (CX 7327G). In its February, 1976 settlement proposal GM mentioned that the elimination "would be likely to increase competition" as "all 12 000 GM dealer locations would be able to claim wholesale compensation whenever they sold an eligible crash part. . . to an independent auto body repair shop (ALJX 13G, H, Supp. to CX 7012) (CCPF 257). 108. GM pays wholesale compensation on sales by one GM dealer to another in two situations. The first is when the (24Jpurchasing dealer is a so-alled Ucountry dealer -a Hsmall dealer in an outlying area" (CX 7813B) who, with advance approval from GM, buys from another GM dealer of the same franchise (CX 7813B; Tr. 2087). The second situation is sales non an emergency basis" when a vehicle is "inoperative" (CX 7813B; Tr. 2087) (CCPF 31I). 109. The "great majority" of collsion damaged vehicles that are repaired can be driven, however emergency basis" orders occur frequently, resulting in a considerable number of wholesale compensation payments (Tr. 1576). A purchasing dealer need only state that his purchase is for an emergency for the selling dealer to justify a claim for wholesale compensation (Tr. 10573-74; 1I063; 12024-25). 1I0. GM dealers consider an "emergency basis" to exist when 464 Initial Decision damaged vehicle is in the shop of the dealer buying the part regardless of whether the vehicle to be repaired is in fact inoperative (Tr. 9054-55; 11062 3). Neither the selling dealer nor a GMPD field warehouse has the means of verifying whether a sale of crash parts to another dealer is in fact for an "emergency" (Tr. 10357-58; 12024- 25). Generally, no attempt at verification is made by the sellng dealer (Tr. 11063-4). The purchasing GM dealer alone determines whether there is an emergency (Tr. 2087) (CCPF 312). 111. Between 1972 and 1976, in auditing the wholesale compensation claims of less than 5% of its approximately 12 000 dealers GM recovered $1 664 194 in erroneous claims (CX 7229). Most of this recovery was due to ineligible sales between GM franchise dealers (CX 7230A, B) (CCFF 313) (RRB 313).

112. GM offcials recognize the problem of possible abuse of wholesale compensation on dealer to dealer sales. These offcials view wholesale compensation on crash parts as faciltating cheating on the claims and have sought to eliminate payments on sales between GM dealers (CX 7346A). GM's president has expressed concern over the amount of wholesale compensation claimed on dealer to dealer sales in some areas, and in general over the significant amount paid in total on such dealer to dealer sales (CX 7253D-E in camera) (CCPF 314).

113. Inventorying of all crash parts at the dealer level would be crashuneconomical because sixty-five percent (65%) of GMPD's parts sell at a rate of 500 or less each year. Ninety-six percent (96%) of crash parts have sales of fewer than 5 000 units per year or less than one per dealer (Tr. 10084) (RPF 55). 114. In the 1976 hearings before the Senate Commerce Committee, a representative of ASC testified that for then current GM models, after the first six months, and for models up to three years old, parts were readily available in a day or two (ALJX 17, pp. 63-4) (RPF 59). (25) 115. GM dealers have an interest in the repeat purchase of cars and therefore an incentive to insure the availabilty of parts, even though a particular customer might make his repeat purchase from another dealer (Tr. 15755-56; 10069-70; 8622-23). GM dealers or their employees testified that their new car sales are affected by the availability of crash parts (Tr. 10475-77; 10820-22; 11841). 116. In all but a few remote areas of some counties in the United States, there are two or more dealers in each car line in the sale of GM crash part. The coverage is such that at least two dealers are within an hour s drive of nearly every populated area in the United States (RX 33A-F) (RPF 65).

Initial Decision 99 F. 117. Some GM dealers selling crash parts over wide areas offer same-day or next-day service within a radius of several hundred miles and impose no minimum order requirement. Some absorb freight costs for orders over $200-$300 (frequently medium or average size orders) and accept returns without penalty, including absorbtion of return freight unless the customer is at fault (Tr. 10835-39, 10844; Tr. 10482-84; Tr. 5527-28) (RPF 70). 118. Approximately five thousand invoices and other evidence of sales issued by 82 GM dealers to the IBS witnesses and the stipulated summaries thereof, indicate that the average price at which these IBSs purchase service GM crash parts was list minus 27% in 1974 and list minus 28% in both 1975 and 1976 (CX 2; CX 5373; Revised CX 5374 - Second Revision CX 5706). By individual trade area, crash parts were purchased at the following discounts off list for the following years:

AVERAGE DISCOUNT ('Y) TRADE AREA 1974 1975 1976 Buffalo, N.Y. Mansfield, Ohio Cleveland, Ohio 26 New Orleans, La. 27 Spokane,St. Louis, Mo.Wash. 26 Tucson, Ariz. 28 (Second Revision CX 5706) (CCPF 107). (26) 119. Dealers generally sell crash parts for which they do not receive wholesale compensation at a price of list less 25% (Tr. 140I9 et seq. 14539-40). They buy these parts at "dealer price, which is list less 40%. Thus, for a noncompensable crash part listing for $100 the dealer buys it for $60 and sells it for $75. The gross profit of $15 divided by the $75 sellng price gives him a 20% gross profit margin (Tr. 14022-26) (RPF 82).

120. Dealers receive a stock order allowance of 5% of "dealer price" on all orders placed on bi-monthly stock or PAD orders (Tr. 11548-9). When a dealer buys a noncompensable crash part "on the pad", his cost is list less 40% less 5%. For example on a $100 list part, his cost is $60 less 5% of $60 ($3.00) or $57. If he sold the part for $75, his margin would then be $18 ($75 less $57) divided by $75 or 25% above his cost (RPF 83).

121. If a GM dealer wholesales noncompensated new GM crash parts at suggested trade prices, he realizes a profit margin of 20% ).

464 Initial Decision and a mark-up of 25%. If the parts were purchased on stock order (PAD), the gross profit margin would increase to 24% and the markup to 31.6% (Tr. 15064, 15067-68) (CCPF 51). 122. Complaint counsel' s expert witness, Dr. Steven Nelson estimates that approximately 50% of crash part orders are on the PAD and subject to the 5% stock order allowance (Tr. 14532). By weighting the margins by the percentage of purchases on and off PAD, dealers' gross profit margin on non- compensable crash parts is 22% (RPF 84).

123. Considering wholesale ' compensation as a reduction in cost rather than as part of the amount realized on the sale of the part, on PAD orders for an eligible crash part dealers pay 40% of list, less wholesale compensation of 300/0 of Hdealer price, less 50/0 of "dealer price . Thus, for a $100 list part, the dealer s cost becomes $42 $100 less $40 less $18 for parts not on the PAD, and $3.00 less or $39 for parts on the PAD (RPF 85).

124. Dr. Nelson estimated that compensable crash parts are generally sold by dealers at prices ranging from list less 25% to list less 40% (Tr. 14019). By taking this range of selling prices, dealer margins on compensable crash parts, again with the 50/50 PAD ratio, may be calculated: (27) Margin Margin, Margin lin Price No Pad Pad Order 50/50 Pad list less 25% ($75) 44% 48% 46% list less 40% ($60) 30% 35% 32. (RPF 86) 125. Dr. Nelson testified that approximately 60% of crash parts sold by dealers are compensable and 40% are noncompensable (Tr. 14535 et seq. Factoring the margins for compensable and noncompensable crash parts in this ratio yields the following dealer gross profit margins for all crash parts wholesaled: Sellng Price Selling Price Noncompensable Compensable Parts Parts Calculation Margin list less 25% list less 25% 6(46%) -I .4(22%) = 36.4% list less 25% list less 40% 6(32.5%)+.4(22%) = 28. Therefore, if wholesale compensation is treated as a reduction in the dealer s cost of the part, dealer margins range from 28.3% to 36.4% (RPF 87).

126. Again using the example of a $100 list part, if the dealer Initial Decision 99 F. stocks the part he would have $60 tied up in it while it is in inventory. After the dealer sells the part, whether it is the same day he purchased it or two years later. he receives, under the example $78. In terms of the amount of money tied up in inventory, his gross profiabilty in wholesaling crash part, and the desirabilty of wholesaling crash part in relationship to returns available elsewhere upon the investment of $60, it makes no difference to the dealer whether the purchaser of the part hands him the $78 or whether the purchaser hands him $60 and a third party, GM, hands him $18 in wholesale compensation. In either event, he would have had $60 tied up in the part and would not receive the $78 until the part is sold. Therefore, the question is: "How do dealer gross margins compare with margins of other types of wholesalers?" or "How do dealers gross margins compare with margins of other types of wholesalers who say they would need to sell crash parts profitably?" It is only by treating wholesale compensation as part of the income on the sale of the part, that meaningful comparisons can be made (RPF 88).

127. Finally, again relying on Dr. Nelson s testimony that approximately 60% of crash part sold by dealers are compensable and 40% are noncompensable (See Finding 125), the following are the dealer gross profit margins for all crash parts wholesaled: (28) Selling Price Sellng Price Noncompensable Compensable Parts Parts Calculation Mar list less 25% list less 25% plus compo 6(37%)+.4(22%) 31.0% list less 25% list less 40% plus compo 6(25%)+.4(22%) 23. (RPF 92) 128. Dr. Nelson s calculations regarding dealer gross profit margin ranges on crash parts wholesaling are confirmed by actual GM dealer financial data disclosing gross profit margins on dealer total wholesale part business. Actual data demonstrate gross profit margins ranging from 25.8% in 1972 to 23.8% in 1977 (RX 301) (RPF 93).

129. If GM chose to sell to them, the most likely candidates for entry into wholesaling new GM crash parts are IWs already sellng other products such as auto glass, rechromed bumpers, automotive paint, abrasive, body shop supplies, and salvage parts to body shops. Other candidates would be cooperatives formed by body shops (Tr. 13915-22) (CCPF 193) (RRB 193).

130. IWs generally provide at least one, and often more, free same day delivery on the products they currently sell (Tr. 5463, 5475 464 Initial Deision 5489- , 705&-8). IWs believe that they could provide this same service in wholesaling new GM crash part (Tr. 5480, 6414) (CCPF 232).

131. Some GM dealers, including some large wholesalers of crash part, do not provide delivery service or provide poor delivery service. Others provide excellent servce (Tr. 3812-13; 12050-1) (CCPF 232A).

132. Some GM dealer wholesalers of crash parts use a very limited sales force, and have no sales force to call on their crash part wholesale accounts (Tr. 3019, 5467). If they have no sales force orders are solicited and taken by phone (Tr. 2726; 6491) (CCPF 239) (RRB 239).

133. When IWs and GM dealers compete in wholesaling on products such as glass, muffers or AC-Delco part GM dealers fare badly in securing such wholesale business or do not attempt to compete with IWs (Tr. 10876) (CCPF 244).

134. IBSs have formed cooperatives which distribute products other than crash part to their members. For instance Consolidated Automotive Part, Inc. ("CAPI") is a St. Louis cooperative which has been in existence since 1973 (Tr. 2313). It handles a variety of automotive part, including sandpaper, paint and crash part applicable to Porsche, Audi, and (29)Volkswagon automobiles. (Tr. 2314). CAPI marks up these items 15% when resellng them to its members (Tr. 2320). One of the reasons for the founding of CAPI was anticipation that GM crash part would become available to wholesalers such as CAPI (Tr. 2317, 2469, 2471). 135. Many IWs have warehouses, delivery equipment, and personnel that could be used for the storage and delivery of crash parts. For example, IWs currently operate the same types of delivery equipment used by GM dealers who deliver crash part (Tr. 5409 5464). Additional personnel, warehouse space, and/or vehicles would pose no problem for IWs (Tr. 2323-24, 6414, 6522, 13915-17) (CCPF 202).

136. Many of the items which IWs sell to body shops are bulky. For example, glass is as bulky as fenders. (Tr. 7666, 7671-72). Windshields are stored in large racks similar to those used for large crash part (Tr. 11167). IWs also inventory heavy and/or bulky items such as 55 gallon drums of thinner and antifreeze, heavy equipment spray booths, masking paper, rebuilt motors, salvage crash parts exhaust system part, and rechromed bumpers (Tr. 3801, 3823, 5460 6398 6423 6472) (CCPF 206).

137. Several IWs who expressed an interest in entering into the wholesaling of GM crash parts currently stock in excess of 10 000 Initial Deision 99 F.

parts numbers. (Tr. 3803, 5459, 11836). For instance, one wholesaler of mechanical parts, paint, and body shop supplies, inventories over 100 000 part numbers, including some items that "turn" only 1 y, times a year (Tr. 8846, 8893). Wholesalers of paint and related items and rechromed bumpers may stock 8 000-10 000 parts (Tr. 3803 5460). Glass wholesalers stock from 2 000 to 6 000 parts (Tr. 5406 7671-72). CAPI in St. Louis stocks 30 000 parts (Tr. 2322)(CCPF 212). 138. There is little difference in the average speed of movement between the parts stocked by IWs and large GM dealer wholesalers. (Tr. 12022, 12025, 12042). The turnover of crash parts sold by large GM dealer wholesalers is 3.2 to 5 times per year (Tr. 11893, 12042). 139. Interested IWs would be willing to invest "whatever it takes" to get into crash part wholesaling (Tr. 2322, 3050, 4408). For example, a national wholesaler of rechromed bumpers would be wiling to invest $2 y, millon to $5 milion initially to enter crash part wholesaling (Tr. 7800, 7857), 3039); $500 000 to $1 milion (Tr. 3039); $250,000 for initial inventory of crash parts applicable to a single GM car line (Tr. 5473a, 5477) (CCPF 215). 140. IWs have several other incentives to enter into the wholesaling of crash parts. Wholesaling such parts complements their current business (Tr. 4402). IWs could spread their (30Joverhead and reduce unit costs by combining deliveries and using their existing sales force (Tr. 6403- , 6515- , 7694, 8871). Crash Parts and Their Manufacture 141. In 1975, GM gross sales of crash parts were in the hundreds of milions of dollars (CX 7407 A in camera). Based on 1977 wholesale compensation payments and a 60% eligibilty factor (i. e., the 13 000 parts on which wholesale compensation is paid out of a total of 32 000 crash parts (Tr. 10072)), GM's gross sales of crash parts had increased more than 70% by 1977 (CCPF 46 and 306 camera) (CCPF 15).

142. Crash parts are sold to a distinct class of customers, body shops which specialize in the repair of crash-damaged vehicles. These shops generally perform very limited mechanical repairs doing such work only when it is accident related. Consequently, body shops' purchases of automotive replacement parts consist almost entirely of crash parts (Tr. 3059-60, 3835) (CCPF 54). 143. While the storage of some crash parts requires special bins such bins in general are not unique (Tr. 10997, 11046-7). Many such bins are built to handle a particular type of crash part (Tr. 11046- 464 Initial Decision 48). Others can be used to store either crash parts or mechanical parts (Tr. 11047) (CCPF 204).

144. Normally GM crash parts wil not fit or coordinate with vehicles assembled by companies other than GM, and crash parts applicable to non-GM vehicles wil not fit or coordinate with GM vehicles (Tr. 1192-94, 1365-6, 1678) (CCPF 56). Crash parts are not standard and usually may be used only for the vehicle for which they were designed (ALJX 9P - Supp. to CX 7006B). 145. Replacement of parts for reasons other than crash damage such as rust, is infrequent. Also, unlike replacement mechanical or functional" parts applicable to GM vehicles, crash parts are seldom installed for purposes of maintenance or due to wear or mechanical failure (Tr. 1361-62, 1675 2260-1; CX 7226B) (CCPF 58). 146. Crash parts account for approximately 70%, both in units and dollars, of all automotive parts replaced under insurance claims for damage to GM automobiles and light trucks (CX 7405; see also CX 7400A- , CX 7401A-X and Z-3 to Z-36, CX 7402A-H) (CCPF 59). 147. Some GM crash parts can seldom, if ever, be repaired due to their type of construction. For example, parts made of pot metal such as fender and rear end caps, parts which are chrome-plated such as bezels, griles, moldings, and glass components generally are not repairable (Tr. 2501, 3068, 3837) (31j(CCPF 62). Parts constructed of plastic, fiberglass, and aluminum are very difficult to repair (Tr. 1729- , 2501--2) (CCPF 64).

148. Dr. George Benston, GM's expert witness, testified that ownership of the dies used in the manufacture of new cars is the decisive, competitive advantage that GM has over other potential manufacturers of crash parts (Tr. 15747) (RPF 180). Usually there exists a single set of dies which produces both original equipment and service parts. The total cost of dies used in the manufacture of crash parts for a particular vehicle varies greatly. It depends upon the number of plastic rather than metal parts, the number of body styles covered and the amount of use obtainable from tooling for prior model vehicles (ALJX 9S, Supp. to CX 7006). 149. A new model normally utilzes substantial tooling from the prior model which may be completely unchanged or modified to create new lines through the use of inserts. The cost of tooling to manufacture crash parts can run to tens of milions of dollars. This cost stems from producing original equipment as well as service and crash parts for repairs. GM estimates that service parts normally account for less than 15% of the total volume of dollars spent on tooling (ALJX 9T, Supp. to CX 7006).

150. Original equipment and repair parts are often produced Initial Deision 99 F.

during the same run. They also may be produced on a separate production run, but within the same plant. For these, setup costs (excluding transportation charges) may run to several thousand dollars depending upon the number of dies required for each part and the complexity involved (e. 17 separate dies were required for an outer rear quarter panel on the 1976 Chevrolet Impala) (AUX , Supp. to CX 7006).

151. No evidence was adduced indicating that GM has impeded entry into the manufacture of parts, including crash parts (AUX , Supp. to CX 7006). Any manufacturer who cares to is free to make GM crash part (CX 7008).

152. "Total crash part demand is high, . . . but with possible exceptions, the demand for each individual part is probably quite modest. The probability that a car wil require the replacement of a particular fender or other crash part with a new one during its lifetime is not great. One would not, therefore, expect this market to be attractive to potential entrants." (CX 7006E, quoting the Commission s Offce of Policy Planning and Evaluation (OPPE), 1975 Semi- Annual Budget Review, Jan. 24, 1975 at p. IW-17). 153. For anyone of the thousands of individual GM crash part the demand is extremely low-particularly when compared to the number of car models originally produced. For example, (32Jbetween 1968 and 1975, fenders for Chevrolets alone accounted for about 11 of production. The following table, included in a March, 1976, GM presentation at the Congressional hearing mentioned above (Finding 54) shows this in greater detail:

Annual 3926848 Production 3926836 3953693 3953694 Front Calendar Year Hood Fender L/H Fender R/H Bumper Bar 1968'" 2.711 784 854 7.431 1969 17, 120 25,097 331 937 1970 141 371 559 31,048 1971 13.917 637 21,688 228 1972 712 17,651 435 11.035 1973 296 397 675 149 1974 544 703 405 975 1975 755 614 211 675 Total 196 119,254 122 158 118.478 % of Production 6.4% 10. 11. 10. (AUX 14Z24, Supp. to CX 7014).

154. These figures ilustrate one of the principal reasons why other manufacturers have not entered the business of manufactur- _., g., 164 Initial Decision ing crash parts. The replacement parts represented by the table above were built as demand warranted during the lifetime of the vehicle. Neither GM nor anyone else could maintain enough warehouse space to economically produce anticipated crash parts needs, such as those involving these four parts, in one production run. Thus, over the years the dies for these units must be set up and the parts produced as inventory needs demand and warehouse space allows. This adds to consumer costs (AWX 14Z24, Supp. to CX 7014). 155. The relatively low demand for crash parts is only one factor which has discouraged other manufacturers from making them. There also are the tooling costs. There are economies if the dies that are required for most crash part are used for producing both original equipment and service parts and such (33Jeconomies are greater when original and servce parts are produced during the same production run. Some crash parts have little or no year-to-year variation in their design. Thus, parts of several years past can be scheduled along with current original equipment production. For example, the trunk assembly for the Chevrolet Monte Carlo used some of the same tooling for model years 1973 through 1976 (AWX 14X, Supp. to CX 7014).

156. GM's Fisher Body manufacturing plants retain dies to make sheet metal parts for models six to seven years old-and in some cases even older. Typically, these dies, which may weigh ten tons and over, are kept in storage yards. The dies are retrieved, steam cleaned, reconditioned-in some cases partially rebuilt-and then inserted into presses to run the required supply of service parts. This is an expensive process because much of it is short-run and labor intensive (AWX 14Y, Supp. to CX 7014).

How Parts Are Distributed By Gm 157. Parts for GM's vehicles are distributed by General Motors Parts Division (GMPD) and AC-Delco Division. GMPD and AC-Delco are engaged wholly in warehousing, marketing, distributing, and sellng parts for GM (Tr. 1994-95).

158. Some parts, such as spark plugs, shock absorbers, radiators, oil fiters, fuel pumps, etc., are sold both by GMPD and AC-Delco. Sheet metal parts, which generally includes crash parts body frame, chassis parts, interior trim parts, and engine parts are sold exclusively by GMPD (Tr. 2003-4). Batteries are sold exclusively AC-Delco (Tr. 2011).

Initial Decision 99 F. GMPD 159. GMPD was established to provide GM car dealers with the parts they need for the make of car they sell (Tr. 1995). GM' s profit from distribution of parts had declined due to rises in warehousing and distribution expenses. These had doubled between 1962 and 1968 for the five car divisions (CX 7248B). The number of service parts needed to serve the market had increased from 132 000 in 1955 to 316 000 in 1968 (CX 7248C).

160. Within GM, GMPD is responsible for assuring the availabilty of part to service GM cars. Aside from batteries, GMPD sells all parts applicable to GM cars (Tr. 10043-5, 10181, 2012) (RPF 17). 161. From model year 1959 through model year 1970, each car division either manufactured or purchased all crash parts applicable to its make of cars. As of January 1, 1959, the predecessor of the present GMPD, which at that time was a part of Chevrolet Division was made responsible for warehousing and distributing servce part to Chevrolet, Oldsmobile, and Pontiac (34Jdealers. Both Buick and Cadilac had assumed responsibility for the distribution of their own service parts, and maintained their own field warehousing operations. Buick and Cadilac field locations were consolidated into GMPD between 1963 and 1966. However, each car division continued to operate a factory warehouse supplementing the GMPD field distribution centers. Many slow-moving parts, including some crash parts, could be obtained only from those warehouses (CX 7002F). 162. GMPD was made a separate division of GM effective March , 1969, and began to assume all procurement and warehousing functions, both field and factory. On September 1, 1970, GMPD became fully responsible for the procurement and warehousing functions. Initially it was not uncommon for GMPD to obtain current model servce parts from Fisher Body Division, although Fisher had itself purchased the parts from an outside supplier. By the beginning of the 1973 model year the transition to GMPD had been concluded (CX 7002F).

163. Approximately 65% of the crash parts which GMPD sells are manufactured by alled GM divisions (e. Chevrolet, Pontiac, Oldsmobile, Buick, Cadilac, Fisher Body) (CX 7011B). 164. Before GMPD was formed each GM car division had its own warehouse parts plant (Tr. 2039; CX 7248D). At such parts plants additional processing was and stil is done on the part before shipment to a dealer cleaning, finishing, painting, protective material applied or packaged (Tr. 2039; 10046-7) (RPF 22). 165. Crash (sheet metal) parts, chassis parts, interior trim parts 464 Initial Decision and engine parts are sold by GM exclusively through GMPD. In addition, parts with GM applications that are sold through AC-Deleo are also sold through GMPD (Tr. 2003, 10068, 10271) (RPF 18). 166. GMPD employs about 13 000 persons. The division does not operate manufacturing plants, but buys parts in a finished state from about 2 500 suppliers-both within General Motors and from outside suppliers (ALJX 14Z-21 , Supp. to CX 7014). 167. GMPD distributes about 300 000 parts. Of that number 000 are AC-Delco parts with GM applications (Tr. 10271), and about 32 000 are crash parts as defined in the Complaint (Tr. 2209). The remainder, about 256 000 part are neither crash parts, nor AC- Deleo parts (Tr. 14678, 10178) (RPF 19). GMPD makes no distinction between crash part and other parts (Tr. 10062) (RPF 20). 168. Of the 32 000 crash parts, 13 000 are eligible for wholesale compensation (Tr. 10072). These 13 000 account for an estimated 60% of the dollar volume of sales of GM crash parts (Tr. 14535-36 and see RX 311A) (CCPF 46). (35) 169. In 1974-75, GMPD conducted a study of crash parts covering 155 separate part numbers. It showed that 23% of the part numbers account for 87% of the sales. The balance, 77% of the total of crash parts, or over 10,000 parts, had sales of less than 700 units a year. The fastest-moving, those with sales of from 600 to 699 units a year, had average dollar sales of less than $5 000 in the twelve months ended May 31, 1975, and an annual rate of inventory turnover of less than 1. Most of the 13 155 part numbers had annual sales of less than 300 units and $2 000 with an annual inventory turnover rate of less than .5. Four percent (4%) of the total, or 537 part numbers had piece sales as high as 5 000 during the year. These sales averaged as high as $40 000 (CX 7006D). 170. GMPD only sells replacement parts to GM passenger car and truck dealers (Tr. 1994). The dealers sell the parts through their own service operation to car owners, to independent body shops, and to other GM dealers (Tr. 1994-95).

171. Currently, GMPD maintains seven factory warehouses part plants") and 36 field warehouses (PDCs). Faster moviIig parts are shipped to GMPD field warehouses. The slowest moving are held at the parts plants to be shipped directly to GM dealers as ordered (Tr. 2039-40 10046-7 10107) (CCPF 17; RPF 24). 172. There are 25 GMPD field warehouses which stock parts for GM automobiles and light trucks (Tr. 10047). These 25 are located in or near major cities of the United States (RX 19C, D). For purposes of placing and receiving orders, the approximately 12 000 GM dealers are assigned to one of the 25 field warehouses, also referred to as g., Initial Decision 99 F. PDCs or parts distribution centers (Ans. n 11; Tr. 2040 2046; 10047- 48) (CCPF 18).

173. This channel of distribution is known as the independent aftermarket (Tr. 12537). There are about 1 000 to 1 500 WDs and over 30 000 jobbers (Tr. 12539; 8640-1). Parts sold in the independent aftermarket are replacement parts, that is, parts that tend to wear and are replaced periodically (Tr. 12490; 12614-15). Generally, there are two or more manufacturing sources for each product line (Tr. 12538). Firms in the independent aftermarket carry parts for most makes of cars (Tr. 12538;Tr. 12633) (RPF 15). 174. The car-part PDCs are: ten "Z" PDCs which carry the 12 000 fastest moving part; nine 11M" PDCs which carry the 12 000 u parts plus "M" parts, which are the 25 000 next fastest moving parts; six "MF" or Master Factory PDCs carry the 12 000 "Z" parts, the 000 "M" parts, plus the "MF" parts which are the 28 000 next fastest moving parts (Tr. 10048) (RPF 25). (36) 175. Of the 300 000 different parts in the GMPD system, approximately 225 000 or 75% are "F" parts. Of the 32 000 different crash parts, approximately 56% are "F" parts (Tr. 13905). Of the 25 field warehouses, six are "MF", nine are "MOl, and ten are "Z" (RX 19A- Tr. 10045-6) (CCPF 20).

176. The PDCs sell to the GM dealers (RX 19; Tr. 10043, 10045- 46). When a dealer places his order with his assigned PDC, a computer discloses where the part is available and, if necessary, refers the order to the nearest PDC or to a parts plant (Tr. 10090-97 10099-100) (RPF 29).

177. GMPD's order fill rate the percentage of items ordered which are in stock at the point of initial order, was 95% for the 1978 model year (Tr. 10061) (RPF 30). In contrast, a GM dealer rarely can fill an entire order for crash parts from inventory (Tr. 3162, 3272 5956-7).

178. Most IBS complaints involve low-demand, slow-moving parts (Tr. 1866, 6982--3). GM dealers generally stock the fast-moving " parts which constitute about 8% of all GM part numbers, and these dealers only stock the "Z" parts which are applicable to their franchise line (Tr. 10242-44).

179. Many GM dealers rely on the GMPD warehouses or another dealer as the primary source of stock for crash parts rather than carry their.own inventory (Tr. 6520).

180. Part plants stock the bulk of all parts, including the slowest moving and older parts. In general, the parts plants supply the PDCs, not dealers. However, the dealers are supplied very slow moving parts and some special-order parts from the part plants. For 464 Initial Decision these slow moving parts, which are stocked principally at the parts plants in Michigan, GM uses air shipment. Dealers normally are supplied from the PDCs. The PDCs receive the parts from parts plants and from manufacturers (Tr. 2040). (37) 181. The following shows the geographical locations of these GMPD facilities:

Parts Plants (6) Master PDCs (9) Flint (Complex = Baltimore Plant 01 , 02, 03 Boston Grand Blanc and Cincinnati Toledo) Denver Martinsburg Jacksonvile Drayton Plains Livonia Pontiac Los Angeles Lansing Minneapolis Detroit Portland Zone PDCs (12) Master Factory PDCs (6) Buffalo Atlanta Cleveland Chicago Houston Dallas Indianapolis Englewood (Newark) Kansas City Oakland Louisville 51. Louis New Orleans New York Omaha Philadelphia Pittsburgh Richmond (ALJX 14Z-22, Supp. to CX 7014) (38) :

Initial Decision 99 F. 182. The following diagram shows the system more graphically (RX19).

GENERAL MOTORS PARTS OIVISION PARTS AND ACCESSORIES DISTRIBUTION SYSTEM 32 ALLIED DIVISIONS & SUBSI DIARIES 65% OF VOLUME MANunCTURIN6 SOURCES 26B OUISIOE SUPPLIERS 300 . VOLUME 6 MAJOR LOCATONS IN MICHIGAN PARTS INCLUOING . TRUCK PLAN I PLANIS 300 000 PARI NUMBERS 12. 5 MILLION. SQUARE FEET 1 UJGATION IN MARTINSBURS . W . VA.

225 000 F PT. NO' 3E PARTS DISTRI BUTION 2 "' Llion CENIERS SIRAIEGICALLY POG' SQUARE FEET LOCATED 464 Initial Decision AC-Delco 183. AC-Delco maintains seven field warehouses and stocks 000 part numbers consisting of over 30 product lines. These lines include spark plugs, fiters, carburetors, fuel pumps, wire and cables seal beam units, shock absorbers, and ignition parts. AC-Delco sells to approximately 3 000 customers. These consist of WDs who sell to jobbers and occasionally to the customers of jobbers such as independent garages, gasoline stations, and car dealers (Tr. 1995-96 2003 2050-1 2210) (CCPF 10).

184. AC-Delco also sells to national accounts such as Firestone Tire & Rubber Co. and Montgomery Ward & Co. (Tr. 1995). Some WDs make some sales to GM dealers (Tr. 1996), however, GM dealers are not permitted to buy part directly from AC-Delco (Tr. 2001). 185. AC-Delco part are generally items that can be used on both GM and non-GM cars (Tr. 2015. , 2230-32). Basically they are parts that are required in the maintenance of the car spark plugs filters, shock absorbers, points, condensers, bulbs, headlamps, fuel and water pumps. (Tr. 10183).

186. With few exceptions, crash parts, as defined in the complaint, are not AC-Delco parts (Tr. 2003-4) (RPF 13). Chrysler's Mopar Parts Distribution Experience 187. Before the early sixties, Chrysler sold its parts from five or six Chrysler-owned plant warehouses to 10 (in 1963) independently owned Mopar wholesalers who resold the parts to Chrysler dealers and other retailers (RX 21Z-0) (RPF 111). In the early sixties Chrysler began phasing out the Mopar wholesalers and replaced them with 13 or 14 Chrysler-owned field warehouses. It also divided its parts into two groups and began sellng them in two separate channels. All parts applicable to Chrysler cars were sold directly to Chrysler dealers. Single source parts, including crash parts, were sold to the dealers exclusively. Parts not applicable to Chrysler cars were not sold to the dealers. In other words Chrysler switched to a system like GM's (Tr. 8486-8, 8494) (RPF 112). 188. The system that Chrysler discarded is similar to the one which the order proposed in the Complaint would establish. Chrysler spent $53 milion to change to a system like GM's. Chrysler has not chosen to switch back. Chrysler s choosing not to switch back is significant because as Dr. Benston, GM's economist witness, noted Chrysler is: (39) now in a position of selling off asts, of contracting their operations. . The simple Initial Decision 99 F. thing, it would seem to me, for them to do would be to sell off this whole system, sell off the warehouses, disband the system, go back to the old system and have a better way of servng consumers for their own benefit or saving resources or something else. They. are selling off a lot of things. They are not sellng off their warehouses, they are not shifting to independents, to my knowledge. I can t think of any better evidence that people who have had previous experience with another system are in a position of wanting to disband some part of their operations, choose not to disband that part of their operations. (Tr. 15818-19) (RPF 119). 189. The $53 millon cost that Chrysler incurred in the 60's is an indication that the costs of changing GM's system would be high. Ineffciencies created by changing the system would add costs which would be passed on, ultimately, no doubt, to the consumer/car owner (Tr. 15734). There also would be a cost to the car owner if parts became less available (Tr. 15735) (RPF 125). Ordering Methods 190. There are four principal ways GM dealers order crash parts from GMPD:

1. The Stock Order (PAD order) is for routine restocking. Each dealer has an assigned day every two weeks on which he can place his stock order (Tr. 10053; 10461-64). Parts are shipped within two days of the receipt of the order (Tr. 10053). Dealers receive a five percent (5%) additional discount on such orders. This is because such orders enable GMPD to fill large quantities per order at convenient times; however, the saving is not translatable into specific cost savings (Tr. 2078-0; Tr. 10072- , 10232- , 10236, 10238-39) (CCPF 22). GM prepays the freight charges on PAD orders (Tr. 2079). Neither a minimum inventory nor a r.ninimum dollar amount or number of units need be handled to qualify for the 5% stocking allowance (Tr. 2080).

2. The Supplemental Stock Order (SSO). These orders may be placed at any time (Tr. 10054). Parts ordered in the morning are to be shipped out the next day. Parts ordered in the afternoon are to be shipped on the second day after the day of the order (Tr. 10054 10108). No minimum order is required (Tr. 10056, 10357-58). GM prepays the freight on such orders (Tr. 2081) (CCPF 24). 3. Car Inoperative Order (CIO). This order is used when a car is inoperative because of a lack of parts. GM prepays the (40)freight (Tr. 2081). A CIO order has priority at GMPD ahead of Stock Orders and Supplementary Stock Orders (Tr. 1005&-57). 4. Very Important Part (VIP). The VIP order receives top priority, may include a search of all warehouses and going to the 464 Initial Decision manufacturer of the part. The dealer pays the freight (Tr. 2076-77 1082-3 10058-0 10109-11) (RPF 31).

191. CIO and VIP orders may be used by any GM dealer at any time, regardless of whether that dealer stocks or not, to obtain servce GM crash part for use in the dealer s own body shop or for resale to IBSs (Tr. 11911- , 12073-74). A GM dealer may rely solely on CIO and VIP orders to obtain service GM crash parts (Tr. 12074) (CCPF 228); however, GM imposes a surcharge of two dollars plus 5% of each line item price on such orders (Tr. 10057, 10109- , 11264- 66) (RRB 228).

192. GM part specialists hold that a well run parts department will order part on the following basis: 80% PAD, 15% SSO, 5% CIO or VIP (CX 7332C). GM offcials have stated that dealers near a field warehouse do not order enough part on their bi-monthly stock orders (PAD) because they can get the part immediately on a CIO basis. In short, it simply does "not pay a dealer near a warehouse to stock crash part" (CX 7332C).

193. Orders are filled at GMPD warehouses in the sequence of VIP and "wil call" first, followed by C10, SSO, and PAD in that order (CCPF 22-25; Tr. 8009). GM offcials have stated "GMPD needs more PAD and less CIO and "wholesale compensation is being abused" in that it is being paid to GM dealers with little stocking done in return by the dealers (CX 7355B). 194. Some warehouses limit "wil calls" to certain hours (Tr. 2084-5). GMPD's Baltimore warehouse, to which 377 GM dealers are assigned, can handle up to 60 dealer "wil calls" a day. This is so even though dealers seldom pick up during evening hours (Tr. 2084- , 10398-99, 10401, 10411). The purpose of "will call" orders is to satisfy dealers urgent requirements or, in some instances, to permit direct delivery to a wholesale customer (CX 7238E) (CCPF 26); however, most GM dealers prefer to receive shipment by common carrier, with freight prepaid by GM (Tr. 10423-25) (RRB 26). 195. At the Baltimore warehouse, a master PDC, PAD orders account for 46% of the orders, CIOs 3G-31 %, SSOs 14%, VIPs 1 Y.% and wil calls 7% (Tr. 10436-37) (CCPF 28 & 229) (RRB 229). 196. Except for VIP orders and wil calls, GMPD generally ships part to dealers in trucks. Truck shipments occasionally take over 24 hours to reach some dealers, although not those dealers located within the metropolitan area of the warehouse. If the shipment is from a warehouse to which the dealer is not assigned, transit time may exceed two days (Tr. 10097-99; 10408) (41) Initial Decision 99 F. OM Parts Pricing and Monitoring 197. GM suggests to its dealers a wholesale price of list less 40%, for compensable crash parts and for noncompensable parts list less 25% (Tr. 10071- 10253-54) (RPF 35).

198. GM has not and does not control or monitor: (1) the price dealers charge wholesale customers for crash parts; (2) the territory in which they sell; or (3) the types or classes of customers to which they sell. GM dealers do not have any exclusive right to wholesale crash parts in their franchise line in particular territories (Tr. 2090 1066!H7, 11022-25) (RPF 37).

199. GMPD, in submitting recommended prices for crash parts to the GM officials responsible for the decision, includes comparisons with the prices of similar part for competing manufacturers vehicles (Tr. 10291; CX 7228A-H). Witness Daly, who was employed by Chrysler, also considered competitive manufacturers' prices in pricing Chrysler crash parts (Tr. 8651) (RPF 52). Insurance Companies and Crash Damaged Vehicles 200. Approximately 90% to 95% of all business done by both IBS and Dealer Auto Body Repair Shops ("DBS") is paid for by insurance companies (Tr. 1872, 2399) (INPF 1). This is not likely to decline due to the growing number of state mandatory insurance laws. As a consequence, obtaining insurance-paid business is crucial to body shops (Tr. 2297, 3857) (CCPF 127). For almost all purchases of crash parts the real consumers ultimately are insurance companies (Tr. 1872) (INPF 20).

201. Since 1970, the major casualty insurance companies have substituted in most instances for the two or three appraisal system a system of cost control in obtaining estimates in connection with their paying for repairs of crash damaged vehicles (Tr. 4309-11). Under the two or three appraisal system IBS and DBS auto body repair shop personnel make the estimate and arrive at their own prices for insurance-paid business (Tr. 4314) (INPF 2 & 3). Under the cost control system insurance company appraisers and/or drive-in appraisal centers operated by the insurance company are used to prepare the estimate and arrive at the price. This gives insurance companies more control over the prices they pay for the repair crash damaged vehicles (Tr. 4315) (INPF 5 & 6). 202. Primarily in rural areas, where drive-in claim centers do not exist and on-site appraisals by company or independent appraisers are inconvenient, some companies still operate on a competitive bid system. Under this system, the customer secures (42Jseveral esti- , 464 Initial Decision mates (usually 2 or 3) from body shops and if he does not have a preference for a particular shop, quality of work considered, the company wil then refer him to the body shop with the lowest estimate (Tr. 1575-76, 1682) (CCPF 129).

203. In writing appraisals for the repair of crash damage insurance companies: (1) use standard Hcrash manuals" to determine the time to be allowed to repair the vehicles (Tr. 1439); (2) use the prevailing" or "going" labor rate in the area (Tr. 1450-51) and; (3) use the " prevailing" or Hgoing" discount in the area on crash parts (Tr. 1451, 1452, 1453). Normally, these are determined by the insurance companies (Tr. 5169, 2410, 7624) (INPF 6-9). 204. For claims settled directly with the insured, appraisers for most insurance companies, whether they are at a drive-in claim center or in the field, wil normally calculate the estimate using parts discounts extended by body shops in the area (Tr. 1218, 1319) (CCPF 130).

205. Most insurance companies, including the largest ones such as State Farm, Allstate, Farmers Group, Safeco, Liberty Mutual Nationwide, and Grange, designate certain body shops as "preferred" one-stop" or "competitive" (RX 288; Tr. 1219, 4844, 5797). Such shops generally are those which have agreed in advance with the insurance company to accept jobs at an agreed-upon parts discount and, sometimes, labor rate. A preferred shop wil normally accept the insurance company s estimate without first seeing the vehicle and preparing its own estimate (Tr. 1221, 1322, 1324) (CCPF 132).

206. Usually neither an IBS nor DBS will have insurance-related work referred unless the shop management has agreed to do that work both at the "prevailng" labor rate and the "prevailing discount rate (Tr. 1450-51 , 1452, 1453, 1836-1) (INPF 13). 207. A very sizeable portion (80%) of insurance-paid work on crash damaged vehicles is performed by body shops to which the claimant is referred by the insurance company (Tr. 6862 3). 208. In St. Louis, Cleveland, and Tucson, virtually all IBSs and DBSs extend a parts discount on insurance work. In these areas, GM dealers offer a discount of at least 10%, and frequently more, up to as much as 25% on bumpers (Tr. 1843 8277) (CCPF 138). 209. In New Orleans, Buffalo, and Spokane, only some IBSs give a 10% parts discount while all GM dealer body shops extend discounts of at least 10%, and as much as 20% to those insurance companies that refer the most business (Tr. 3117- , 3156, 3651- 3657, 7363-4, 7915). In Mansfield, Ohio, discounts are given by GM dealer and independent body shops only as included (43)in the initial Decision 99 F. estimates prepared by adjusters in referral situations. However, GM dealer body shops successfully attract most of the insurance business by submitting lower bids (Tr. 6595-96, 6610-11) (CCPF I39). 210. Failure to match GM dealer discounts results in IBSs losing 3655,insurance-paid business to GM dealers (Tr. 1325, 2301--4, 3660-1) (CCPF 141) (lNRB 141).

211. In some instances, IBSs match the discounts offered by GM dealer shops to obtain insurance company referrals. If they do, the IBSs get their fair share of referral work (Tr. 6675-76) (CCPF 142). 212. Some IBSs could not continue in business if they were to meet the discounts GM dealer shops give to insurance companies (Tr. 2851, 3653-54, 4212). AI's Body and Fender Repair in Spokane Washington, for example, used to give a 10% discount to insurance companies but soon found it had Htoo slim a margin" and so couldn afford to continue it. Today AI's gives no part discount but in so doing loses considerable volume. Because he competes with GM dealers who give 15-20% part discounts to insurance companies AI's gets no GM referral work. Consequently only about 10% of his business is on GM vehicles (Tr. 7363-6) (CCPF 145). Repair or Replace with New or Used Parts 213. When a motor vehicle is crash damaged the owner has several options. A new replacement part may be installed or the damaged part may be repaired or replaced with a used part. Some parts may be replaced with a partial panel not manufactured by the new vehicle supplier. Also, the repair may not be made or, if the damage is extensive, the vehicle may be scrapped. If it is, additional used parts becoroe available (CX 7006B).

214. Salvage crash parts are used crash part obtained from wrecked vehicles (Tr. 1385, 1731). Salvage yards purchase wrecked GM vehicles, disassemble them and wholesale the salvageable GM crash parts to body shops (Tr. 1908-10, 4415) (CCPF 68). 215. A survey was conducted by GM in 1974 of thirty-one automobile body repair shops in Fort Wayne, Indiana. The operators of the shops estimated that in fifty percent of all instances of crash damaged vehicles, original parts were repaired and reused (CX 7006B-(;. The repair rather than replacement of parts on crashdamaged GM vehicles has been decreasing in part due to changes in vehicle construction (Tr. 2501).

216. Recent year GM automobiles and light trucks contain more and more crash part made of fiberglass and aluminum to (44Jreduce weight and thereby meet federal mileage requirements (Tr. 1265-6 464 Initial Decision 1384, 1729-30) (CCPF 65). The repair of fiberglass crash parts poses a health hazard and is, therefore, disfavored by body men (Tr. 2501). 217. Frequently parts are replaced when they could be repaired at lower cost. This is because:

(1) Owners may insist on new parts-specially on newer carseven though a repaired part would be satisfactory; (2) Some repair shops work almost exclusively on a high volume basis and can process more jobs by replacing parts than by devoting the time to repairs-€ven though the cost saving to insurers and consumers could be substantial;

(3) In some cases, body repair shops do not or cannot employ persons possessing the necessary skils to repair rather than replace; (4) The insurance adjuster may not have been sufficiently trained to recognize how part can be repaired and to understand the cost equation of repair versus replacement. Lacking this practical knowledge, he may be unable to obtain the insured's agreement to repair the part rather than replace it (ALJX 14Z-, Supp. to CX 7014).

218. Insurance companies have fostered and continue to foster the installation of used or salvage parts (Tr. 1764, 1246) (RPF 191). Allstate recommends used parts whenever possible to avoid total losses (Tr. 1247). Travelers requires used parts installation, regardless of consumer opposition, whenever possible and economical including on current models (Tr. 1510-11) (RPF 192). Salvage Parts 219. Unlike GM dealers, salvage yards and bumper rechromers tend not to specialize in the sale of parts for only one make of vehicle (Tr. 1243-4, 1400) (CCPF 71).

220. Salvage yard operators and rechromers consider their respective businesses to be separate industries from the wholesaling of new GM crash part (Tr. 1947) (CCPF 73). 221. Far more new GM crash parts than salvage GM crash parts and rechromed GM bumpers are used in replacing damaged portions of GM automobiles and light trucks. On claims paid by leading insurance companies for crash parts applicable to GM cars and light trucks, approximately 75%-90% of the dollar (45Jamount paid and approximately 85% of the units obtained are for new rather than used GM crash parts (CX 7400H- , CX 7401A- , W, CX 7403, CX 7405; Tr. 1243-9, 1399-1400, 1568).

222. Several of the parts most frequently needed to repair crash Initial Decision 99 F. damage, such as exterior moldings, griles, fender and rear end caps bumpers, and lamp assemblies are seldom available as salvage. This is due in part to the difficulty of removing them from wrecks and/or the unacceptable condition in which such units are found on wrecked vehicles (Tr. 1392- , 1938-39). Other salvage crash parts which seldom are utilized include rocker panels, wheel opening panels, and quarter panels (CX 7400H; CX 7401A-V) (CCPF 79). 223. In general, salvage crash parts are only available as part of an assembly rather than as individual parts. In other words, salvage yards generally decline to sell individual crash parts from a front end, rear end or door assembly, preferring to sell the entire unit (Tr. 1240, 1391, 1734-35). A front end assembly usually includes the front hood, lamp assemblies, andbumper, grile, left and right fenders, moldings. A rear end assembly usually includes the left and right quarter panels, trunk lid, rear body panel, lamp assemblies, moldings, and rear bumper. A door assembly wil usually include the door skin, door hardware, and door inner panel (Tr. 1239-40, 1390-92 1734-5).

224. Vehicle owners and body shop operators strongly prefer new over used crash parts (Tr. 1251-52, 1255). Some body shops use salvage crash part and rechromed bumpers only when there are no new crash parts available (Tr. 4506-8) (CCPF 86). 225. Salvage crash parts are often bent, rusted, previously repaired, scratched, or otherwise damaged, and in need of removal of old paint. The outside storage of salvage crash parts alone results in their deterioration. Thus, it is frequently necessary to trim salvage crash parts by cutting off unnecessary material and otherwise to expend extensive labor to refurbish them prior to use (Tr. 1396-97 2267-B8) (CCPF 88).

226. If the price of a salvage crash part approached that of a new crash part, a body shop would buy a new part if it were available (Tr. 2510) (CCPF 93).

227. The price charged for any given salvage crash part is primarily a matter of supply and demand, the salvage yards' cost of acquiring the part, and the condition of the part. Consequently, the price of any particular salvage crash part may fluctuate widely. Prices change frequently and vary from one year to the next. Due to the condition of the part the price may vary on a salvage yard lot for the same model year. Unlike GM crash parts there are no set or published prices or price lists for salvage GM crash parts (Tr. 1259- , 1262, 1925-27, (46)1945, 1948-9) (CCPF 95). 228. Almost without exception the price trend over time is exactly the opposite for new crash parts and salvage crash parts. 464 Initial Decision Prices of new crash parts increase as the vehicles they fit become older, while the prices of salvage crash parts fall rapidly with time (Tr. 1748-9, 2511- , 2514) (CCPF 97).

229. As a consequence of consumer preferences and increasing price disparity over time, salvage GM crash parts are utilized less frequently on the most recent GM vehicles. They are used increasingly as the vehicle becomes older, in part, to avoid "totallng e., not repairing the vehicle. In contrast, new crash parts are utilzed almost entirely on newer models and are used less frequently as the vehicle ages (CX 7401 Z9-Z26, Tr. 1246) (CCPF 100). 230. State Farm, the world's largest automobile casualty insurer (Tr. 1667), has actively encouraged the use of used parts. At all times State Farm stresses consideration of used parts, including salvage assemblies and rechromed humpers (RX 200G, RX 212D, RX 223A). If it costs less to replace with salvage parts and they are available State Farm prefers salvage over new (Tr. 1764, 1792). If there is a suhstantial cost difference between new and used crash parts, State Farm makes a settlement offer based on used part, even though consumer preference is for new, and does so even for current and recent models when used parts are available (Tr. 1791) (RPF 194). 231. Salvage yards have made used GM crash parts more readily available by use of long distance telephone and Telex lines to other yards and published parts locators (e. RX 228; RX 236), enabling them to locate and obtain salvage parts on a national basis if they are not stocked locally. Even current model GM crash parts may be obtained from salvage yards on a 24-hour basis (Tr. 992B-29; RX 138M, RX 273A). The long line system is like a conference call. It has speakers on which one person may speak to everyone else on the line (Tr. 1927).

232. Manuals, such as those published by Mitchell' s and Holland- , assist salvage operators in determining which parts are interchangeable for which vehicles (Tr. 1964-5) (RPF 200). 233. The salvage industry has had some success in persuading body shops to use salvage parts (Tr. 1968-9). 234. Rechromed bumpers also compete with new bumpers, are priced relative to new bumpers, and are easy to install (Tr. 7795-96) (RPF 201).

235. Availability of salvage GM crash parts may vary by model year. After the second model year they are more available (CCPF 98) (Tr. 3603, 1386-7). A salvage yard's most frequent (47Jsales are of parts after the second model year. For example, in 1976, about 91% of witness Arnold' s parts sales were for cars older than two years (Tr. Initial Deision 99 F.

1930-31) (RPF 203). Those needed for the less popular GM models are seldom in good supply (Tr. 1249-50, 1389-90) (CCPF 80). 236. The prices of new and salvage GM crash parts are related and new and salvage GM crash parts are competitive with each other (Tr. 1523, 1861). Usually the list price of salvage parts is expressed as a percentage of the list price of the comparable new part, although the percentage factor may vary with the model year of the vehicle (Tr. 2782, 1950-1). As new part prices escalate, the prices of used part tend to follow (RX 278F, G) (RPF 205). 237. An Aetna study of repairs paid for indicated that 23. (dollars) of GM crash part purchases were for salvage parts (Tr. 1546). A similar State Farm survey disclosed that 28.6% (dollars) were for salvage parts (CX 7400; Tr. 1590-91) and a study by Travelers indicated that the figure was approximately 24% (dollars) (Tr. 1493-94, 1497) (RPF 209-11).

238. Dr. Benston testified that used crash part are one of several substitutes for new crash parts (Tr. 15748, 15775). Also, it is stated in an FTC staff memorandum (1973 Annual Planning Report) that new crash part "are in competition with recycled crash parts and in some cases, with repaired part" (RX 51L) (RPF 213). 239. The dollar volume of used automotive parts sold annually is substantial. In a 1969 study, the U.S. Commerce Department Business and Defense Services Administration, estimated that nationally the used part industry provides $4.5 bilion in replacement parts annually, about one-third the dollar value of replacement parts consumed by the automotive aftermarket. (These parts would be valued at $15 bilion new.) Half of these used parts are consumed by the repair trade (RX 138G, R). Mr. Arnold testified that about 40% of this figure would be crash parts ($0.9 billion), and about 63% of those crash parts would be for GM vehicles ($567 milion) (Tr. 1953-54).

240. GM's Motors Insurance Company subsidiary stresses repair over replacement of damaged crash parts (Tr. 3934) as does GM (ALJX 14Z-3, Z-6 and Z-7, Supp. to CX 7014) (RPF 217). 241. Of Allstate s total losses, almost fifty percent of damaged GM crash parts are repaired rather than replaced (Tr. 1265). State Farm studies indicated that parts are repaired rather than replaced 40% to 50% of the time (Tr. 1730) (RPF 220). 242. State Farm has calculated that 43% of the damaged (48) parts (units) including bumpers and exterior moldings covered by its policies are repaired rather than replaced (CX 7400H). Bumpers are replaced by used bumpers 60% of the time in terms of dollar volume of replaced parts (CX 7400L). By excluding bumpers and exterior 464 Initial Decision moldings from CX 7400H, the extent of repair (units) becomes 58% (RPF 221).

Why Have Prices of Crash Parts Increased? 243. The upward trend in crash parts prices is not solely due to GM' s decisions to raise prices because of factors related to profitability. For example, on March 12, 1976, the 1972 Chevrolet Chevelle front and rear bumper part had a combined list price of $146. The list price of the 1975 Chevelle bumper parts on that date was $417. Ofthe total $271 difference, $244 was due to the addition of bumper part which were required to meet the federal bumper safety standard. More details are shown in the table following: GENERAL MOTORS CORPORATION BUMPER PARTS PRICE COMPARISONS CHEVELLE: 1972 VS. 1975 MODEL Paris List Prices Bumper Pari Description March 12, 1976 Chev. Chevelle Difference 72 Model 75 Model Amount Percent Front Bumper Face Bar 78. 82. 11. 14. Front Bumper Reinforcement 101. 101.00 Front Bumper Energy Absorber 21. 21. Rear Bumper Face Bar 67. 63. 15. 22. Rear Bumper Reinforcement 101.00 101. Rear Bumper Energy Absorber 21. 21. TOTAL 146. 416. 270. 185. (ALJX 14Z17, Supp. to CX 7014) 244. The bumper reinforcement and energy absorbers required on the 1975 Chevelle were not required on the 1972 model, which accounts for much of the increase in price. As federal bumper standards become even more stringent in 1979-1980, the crash parts costs for bumpers probably wil increase further (ALJX 14-H, Supp. to CX 7014). (49) 245. Between September, 1971, and March of 1976, GM servicepart prices increased approximately 37%, while crash parts, taken as a specific category, rose 35%. During approximately this same period, the price of steel mil products, from which many crash parts are made, rose 58% and GM's average hourly labor costs rose 51 %. The cost of paper products, which figure importantly in the packaging and distribution of service parts rose 67%. Rail transpor- Initial Decision 99 F. went up 510/0.tation, another significant cost in the parts business, The composite index of industrial commodities rose 55% (ALJX 14K Supp. to CX 7014).

246. In the mid-70' s there was an advertising campaign and representations were made by insurance company representatives to a congressional committee holding hearings on the cost of crash parts that the cost of replacing all the parts to completely repair a 1973 Chevrolet could range up to $24 000. It was not clear that the $24 000 was for parts and labor. GM responded that individual parts always involve far greater costs than finished consumer goods in unit packaging, stocking, cataloging, inventory expenses, obsolescence, and shipping to provide availability. As with most, if not all manufactured products, the cost of buying and installng the individual parts of a car one at a time would be significantly higher than the cost of a production-line-assembled new car or product (ALJX 14L, Supp. to CX 7014).

Does GM's Method of Selling Crash Parts Discriminate Against IBSs? 247. Most IBSs perform repair work on all makes of cars and light trucks, including foreign-make vehicles. However, due to the large number of GM cars on the road, work done on GM-made vehicles accounts for a significant amount of the potential volume for IBSs and generally for a significant amount of their actual receipts (Tr. 5222-23). Formerly, work on GM vehicles constituted a greater percentage of IBSs' business than it does today (Tr. 2599) (CCPF 123).

248. On purchases of crash parts from GM dealers, IBSs pay an average of approximately list minus 22% on noncompensable and list minus 32% on compensable parts (Tr. 10 520, 14,521). On non- PAD orders, GM dealers pay list minus 40% and on PAD orders list minus 43% for both compensable and noncompensable crash parts (RX 311A; Tr. 14515-16) (CCPF 43).

249. Dr. Nelson testified that these averages show that there is a 15.7% average price differential between what IBSs and GM dealers pay for new service GM crash parts on the basis of his estimate that 60% of the purchases are of compensated parts and that 50% of the GM dealer purchases were on PAD (Tr. 14536, 14560-2). 250. Dr. Benston testified that a comparison between the (50) total cost incurred by each, not just a comparison between the invoice prices paid, is a way to determine whether IBSs are at a cost disadvantage in competing with GM dealers (Tr. 16109). 464 Initial Decision 251. Dr. Nelson testified that a wholesaling GM dealer would not resell at his cost to buy a part, that a dealer who transfers a part from his warehouse to his own body shop has real world costs allocable to the transfer and needs a minumum gloss margin of 20% (translating to 25% mark-up over cost) to stay in business (Tr. 14469 14920-23) (RPF 126).

252. The time it takes for an IBS to receive a crash part differs widely depending on whether the part is in the inventory of the dealership from which the part was ordered. If it is in inventory, it may be as little as one hour or as much as 24 hours until delivery (Tr. 2381, 2451- , 4044, 11131- , 10644-6). If a part is not in the dealer s stock delivery time depends on whether another local GM dealer has the part, as well as the proximity of the nearest GMPD warehouse. In warehouse cities such as Buffalo, St. Louis, Cleveland and New Orleans (RX 19C), it may take three days for IBSs to receive parts ordered by the dealer from the local GMPD warehouse (Tr. 4801--2 7503) (CCPF 114) (RRB 114).

253. The parts departments of GM dealers perform services in wholesaling crash parts to both their own body shops and IBSs. The services performed include ordering, receiving. and maintaining inventory (and obtaining non-stocked parts from other dealers or from GM warehouses), labor in stocking and picking parts from shelves for orders, delivery, order taking and interpretation, biling and record-keeping (Tr. 11018, 11035, 11988-9, 10449- , 10473- 75).

254. Costs associated with the inventorying function include labor to receive, store, pull (remove from stock), and load parts financial costs for investment in inventory. as well as the cost of facilties to house the parts, insurance, and equipment to store and move them. (RPF 128).

255. There also are costs in servicing IBSs such as biling costs the costs of extending credit, including credit checks, uncollectable 10487--8). accounts, and money tied up in receivables (Tr. 10846, There are also vehicle and driver expenses for free delivery, and counter servce expense (Tr. 11221-22; 10663) (RPF 130). 256. GM financial studies indicate that the national average of dealers' parts and accessories departments' direct and allocated total expenses is approximately 25% of the purchase price of all parts. Between the years 1970 and 1975, dealers' parts and accessories departments total expenses ranged from 22.6% to 25.3% of the cost of sales (RX 35;Tr. 11421-22) (RPF 132). (51) 257. For warranty repairs, where GM pays for the part, GM pays ).

Initial Deision 99 F. T. its dealers cost (dealer net) plus 30%. This permits the dealer a profit of roughly 5% (Tr. 11430-31) (RPF 133).

258. NADA studies have disclosed that dealer overhead in parts wholesaling is between 30% and 37.5% of the cost of the parts and that there was a need for GM to increase wholesale compensation (Tr. 8187-88; See also, ALJX 17, p. 95).

259. Mr. Daly, the former Chrysler offcial, said that a 25% to 27% margin is required to cover costs of handling crash part at the wholesale level (Tr. 8667) and that a fair dealer markup over cost would be 25% to 33.3% (Tr. 8684) (RPF 134). 260. GM dealers generally mark up crash parts sold or transferred to their own body shops at 25% over the cost of the parts. This is an accounting practice recommended by GM (Tr. 11420-21 4753). In contrast to the warranty reimbursement of 30%, the recommended transfer price of adding 25% to dealer net for parts moving from the part to the body shop departments does not include a profit (Tr. 11430-31). A GM offcial estimated that 60% to 65% of all GM dealers use the 25% transfer price (Tr. 11424). Some GM dealers use a 30% markup (list minus 22%). (Tr. 10491) (RPF 135). 261. When the average cost of the wholesaling function which dealers perform for themselves, as measured by the GM recommended transfer price (25% of dealer net, no profit), or the warranty reimbursement (30% of dealer net, 5% profit), is added to the average purchase price of the part to the dealer, the total cost of the parts installed by dealers averages from about list less 27% to list less 24%. These figures are obtained in the following manner. In order to determine the average purchase price of a part, the allowance must be taken into account. About 50% of crash part orders are on the PAD (Tr. 14532 et seq. With half of the order at dealer net (list less 40%, or 60% of list), the average purchase price is 58.5% of list on a $100 part dealer cost wil be $60 off PAD and $57 on PAD, the average of these two is $58.50 or 58.5% of list. When 25% is used as the measure of the dealer s wholesaling cost, the average total cost of the part to the dealer is his average purchase price plus 25% of dealer net (58.5% X 25% = 14.6%. 14.6% + 58.5% 73.2% of list price) which translates into about list less 27% as the net cost to the dealer for part he uses in his own body shop. On warranty work, where the reimbursement is 30% of dealer net the net cost to the dealer is 76% of list price which translates into about list less 24%. (58.5% X .39% = 17.5%. 17.5% + 58.5% = 76.5% of list price). Without the PAD allowance, the cost to the dealer is from list less 25% to list less 22% (RPF 139). 262. IBSs do not incur the stocking costs of GM dealers in (52) 464 Initial Deision handling crash parts (Tr. 14016). IBSs testified that they do not provide for themselves wholesaling services or incur the costs related to them (Tr. 7662-3, 7973-0) (RPF 140).

263. GM dealers charge IBSs, 25% to 40% off suggested list price for compensable crash parts (Tr. 10847-48, 10664-5). (RPF 141). IBSs Numbers, Sales Volume and Numbers of Employees 264. IBSs are generally smaller operations than GM dealer body shops, the former often being one or two man shops (CX 7327G; Tr. 4796-97 4866 8292) (CCPF 151).

Growth in Numbers - Government Data 265. For the period 1963 through 1967, U.S. Census Bureau Data show that the number of IBSs (SIC Code 7531) increased by 4 621 (28.5%) (RX 39). During this period, GM did not pay wholesale compensation on crash part sales. Wholesale compensation began in September or October, 1968 (Tr. 2096-97) (RPF 152). 266. Between 1967 and 1972, the number of IBSs reporting to the S. Census Bureau grew by 10 982 (52.7%), while the estimated number of GM dealer body shops declined about 343 (3.6%) (RX 38; RX 39) (RPF 153).

267. According to U.S. Census Bureau data, the number of nondealer repair shops, other than body shops (SIC Codes 7534, 7538 and 7539) increased by 24% between 1967 and 1972, a significantly lower growth than that of the IBSs. These non dealer repair shops are categorizd by the Census Bureau as "General Automotive Repair Shops, and "Other Automotive Repair Shops" (RX 39) (RPF 154). 268. IRS data showing the numbers of IBS proprietorships and partnerships (but not corporations) indicate a growth between 1967 and 1972 of 51 % (RX 314A) and a growth between 1967 and 1976 of 55.3% (RX 318). Between I967 and 1976, GM body shops declined in number by 4.2% (RX 38) (RPF I55).

Growth in Sales - Government Data 269. For the period 1963 through 1967, U.S. Census Bureau data show that sales by IBSs grew by $263 milion (47%) (RX 39) (RPF 157).

270. Between 1967 and 1972, sales by IBSs, according to Census data, increased by nearly $952 milion or 116%, while GM car dealer body shops' sales, according to GM data, increased by approximately $226 milion or 40%. During the same period, according to Census Initial Decision 99 F. data, non-dealer repair shops, other than (53Jbody shops, had a sales increase of 62% (RX 38; RX 39) (RPF 158). 271. According to IRS data, between 1967 and 1976, sales by IBS partnerships and proprietorships grew by 151 %. Unlike the Census data, the IRS data do not include corporations (Tr. 14326). Over the same period GM body shops' sales grew by 92% (RX 317) (RPF 159). 272. The growth in sales by IBSs, including corporations, can be estimated by the process of linking the Census data, showing the growth of sales of IBSs (corporations included over the 1967-1972 period, with the IRS data showing the growth in sales of IBSs (corporations excludecl over the 1972-1976 period (Tr. 16027-34). 273. Between 1967 and 1972, IBS corporations grew by a higher precentage than IBS partnerships and proprietorships (Tr. 14370). Assuming that IBS corporations grew by only the same percentage as the IBS partnerships and proprietorships in the period 1972-1976 it is estimated by using the linking process, that IBSs (corporations included) grew in 1967-1976 by 196%, or by more than twice the percentage (92%) that GM dealer body shops grew over the same period (RX 322) (RPF 160).

Dun and Bradstreet Data 274. Dun and Bradstreet data demonstrate that IBSs have continued to grow in numbers, sales, and employees. From 1972 to 1977 the number of IBSs surveyed by Dun and Bradstreet grew 53% , or from 11 644 to 17 864. The number of GM body shops (according to GM data) declined from 9057 to 9001. Sales receipts for these IBSs grew by 85% or from $894 millon to $1.651 bilion. GM dealer body shops' sales (according to GM data) increased by 53%, or from $571 millon to $979 million. According to the Dun and Bradstreet survey, the number of non-dealer repair shops other than IBSs increased in the same period by 20% , a significantly lower growth rate than that of the IBSs. These non-dealer repair shops also experienced a lower growth in sales when compared to IBSs, 76% compared to 85% for the IBSs. The number of employees of IBSs also grew between 1972 and 1977 from 49,438 to 80 019 (62%), compared to employment growth in non-dealer repair shops, other than IBSs from 235 179 to 244 622 (4%) (RX 38, RX 43) (RPF 162). 275. A Dun and Bradstreet study also reveals that the failure rate for IBSs was 15 per 10 000 in 1977, a decline from 32 per 10 000 in 1972. Only one of the 23 retail lines of business for which Dun and Bradstreet maintains failure rates in the normal course of business experienced a lower failure rate in 1977. Motor vehicle franchise 464 Initial Decision dealers had a failure rate of 20 per 10 000 (RX 303A-B; Tr. 12251) (RPF 164). (54) Telephone Directory Listings 276. Based on telephone directory (Yellow Page) listings, the number of IBSs in Buffalo, Cleveland, New Orleans, St. Louis and Tucson has grown over a ten-year period. Between 1967 and 1977 the growth rate was almost 83%. In St. Louis, the growth rate; was in excess of 26%. In Tucson, the rate was about 105%. During the tenyear period, the body shop growth rate for these five areas was 38%. For Spokane, there are no comparable data for 1967, but in the five. year period between 1972 and 1977, there was an increase of over 30% in the number of IBSs (RX 41) (RPF 166). 277. Based on 1977 Yellow Pages listings, more than 72% of the IBSs in Commission counsel's selected areas, excluding Mansfield Ohio, are located where population has been declining. From 1970 through 1975, Buffalo lost 1.6% of its population; Cleveland, 4.3%; and St. Louis, 1.7%. During the same period, the nation s population rose by 4.8%. The two remaining cities, which showed an increase in population since 1967, and for which ten-year body shop data are available, are New Orleans and Tucson. These cities show an increase in the total number of IBSs of approximately 83% and 105%, respectively, which outstripped the percentage increase in the populations of these areas between 1970 and 1975 (RX 41; RX 42) (RPF 168).

278. Several witnesses testified that listings in the Yellow Pages are a reliable method of analyzing numbers of IBSs entering or leaving the business (Tr. 4269, 7047) (RPF 167). Financial Health 279. Some of the IBSs that grew the most in terms of gross sales declined in profitability, or actually experienced losses (Tr. 4184-5; 4816-20). Others that grew in terms of gross sales in fact cut back on their number of body repair men, biled fewer hours or repaired fewer vehicles (Tr. 4205, 4207, 4227, 6557-58, 6590-91, 6650-52). 280. Not only is the apparent profitability of IBSs very low, but the actual return to the owners is even lower than the profit figures indicate due to the owners' low salaries or withdrawals (Tr. 2292- 2483-4 4010-11 4827 4831 4841-42 4859).

281. In recent years, as the profitability of collision work has declined, many IBSs have entered into other fields, most of which are related to collsion repair work (Tr. 15250-51), such as towing (Tr. Initial Deision 99 F. T. 5692); boat repair (Tr. 6775-76); appraisals (Tr. 4794-95 , 4967); salvage pools (Tr. 7039); fiberglass repairs (Tr. 3125); gasoline sales (Tr. 5691-92); offce furniture and cabinet repair (Tr. 7041 , 9373); design and (55Jinstallation of van interiors (Tr. 7042); salvage vehicle rebuilding (Tr. 5987, 9374); patchwork on rusted out cars (Tr. 4945); mechanical and radiator repair (Tr. 3777); body shop supplies abrasives and heavy equipment wholesaling (Tr. 6163-4); rust proofing (Tr. 3994); and frame straightening (Tr. 4823-24, 4967 7095).

282. In practically all cases, these sideline operations have assumed an increasing share of the IBSs' total volume of business because they are more profitable than collsion repair work (Tr. 3681, 4121 , 4190, 4967-68). In some cases, IBS operators have virtually closed down their automobile collsion repair operations in favor of these other enterprises (Tr. 4041 , 9374) (CPF 169). 283. Other IBSs concentrate on specialized collsion repair work (Tr. 8287-91 (heavy duty trucks and equipment); Tr. 2982 (heavy trucks and buses); Tr. 5569 (heavy equipment); Tr. 7911 (Winnebago campers); Tr. 9374 (vinyl tops and customizing, trucks and trailers)) (CCPF 171).

284. Sublet work from dealers constitutes a significant share of the volume of some IBSs (Tr. 5009-10, 7906, 7911-13). IBSs generally are eager to take on any sublet work they can get (Tr. 485S-59; see official notice Tr. 15251--2) (CCPF 174). 285. If IBSs received price parity in obtaining service GM crash part, IBS witnesses testified they would: (1) expand and modernize their facilities (Tr. 9376); (2) purchase new and more sophisticated equipment (Tr. 4864); (3) hire additional or more skiled employees (Tr. 3127, 3668); (4) raise their employees' salaries or provide fringe benefits such as paid vacations, uniforms, and hospitalization benefits (Tr. 2860, 7570); (5) complete their repair jobs more quickly (Tr. 9376, 3866, 4864); and (6) provide for consumers such incidental services as free undercoating and car washes (Tr. 3127) (CCPF 297). 286. Some IBS witnesses . testified that their body shops were operating at or very near capacity (Tr. 6120, 3210, 2912 ("runs a full house," has plenty of business); 404B-9 4121 (80% capacity, in November 1977, already booked through January 1978); 4882, 4945- , 5102-3 (90% capacity currently, in 1977 was full; 8344-5 2632, 2660 (no trouble getting business; could not take on 10-15% more work without improving facilities and adding men); 6259-60 6319 (had all the work he could handle, had to reject jobs, which went to GM dealers)) (RPF 175).

287. A number of IBS witnesses testified that they added to their 464 Initial Deision capacity during the 1970's by construction of additional buildings and the addition of expensive equipment (Tr. 3100, 3189, 3208 (added 29 bays in 10 years, 15-20 bays added in 1973 with new building), 2291 (added buildings and equipment, 1973-76), 5096 (invested $40 000 in frame equipment), 5334-35 (56)($5 000 lift, $12 000 new doors), 390&-7, 3915-16 (equipment and real estate purchased in 1968, 1974 and 197&-77), 7905-6, 4188-9 (1970-new building and frame rack for $20 000)) (RPF 176).

Relationships Between GM, Its Dealers, and Crash Parts 288. Dr. Nelson testified that GM is dependent on its dealers loyalty and good wil to sell its major product-motor vehicles-and that GM dealers oppose losing their exclusive wholesaling privileges on new service GM crash part, a major parts item with them on which they face limited competition (Tr. 14869, 14879). In confining the distribution of GM crash part to its dealers, GM dealers are dependent on GM not only for the vehicles they sell, but also for the crash part to repair such vehicles (Tr. 14833). 289. Various dealer advisory bodies regularly meet with GM offcials. For the General Motors Dealers Council, GM dealers in each of numerous geographical zones vote within their division (i.e., Chevrolet, Oldsmobile, Pontiac, Buick, Cadilac, and GMC Truck) to elect representatives to Regional Dealer Councils. The Regional Councils in each division elect a representative to the Divisional National Council. Each of these six National Councils elects a Chairman from among its membership. The Councils solicit and receive opinions from dealers regarding GM's policies and meet periodically with GM executives to communicate these views and to make recommendations (CX 7208, CX 7209A- , CX 7210A , CX 7211A- , CX 7212A-D, CX 7213A- , CX 7214A , CX 7215A-B; Tr. 2070, 3311- , 3320-21, 3423-3I, 4604, 8137-38) (CCPF 278) (RRB 278).

290. GM also has other dealer advisory bodies such as its President's Committee and its National Advisory Counsel (Tr. 8139 3437) (CCPF 279).

291. GM franchise dealers also are organized to present their views to GM through NADA (Tr. 3422). About 70% of all GM dealers belong to NADA (Tr. 82I2). NADA has an Industry Relations Committee (IRC) which is composed of new car and truck franchise dealers broken down into "line groups" according to car make. At present, the IRC is comprised of various groups, one of which is the General Motors Line Group. IRC, particularly the GM Line Group, is the offcial voice of NADA's GM dealer members (CPF 268). The . . .

Initial Decision 99 F. chairman of the GM Line Group is a GM dealer who is appointed the president of NADA in consultation with the Industry Relations Committee Chairman. In addition to the chairman, the GM Line Group includes each chairman of the six GM Divisional National Councils (Tr. 3316, 3422, 3431-33) (CCPF 280) (RRB 280). 292. Over the past few years, GM dealers, either individually, through their trade association N ADA or through (57)GM's dealer councils, have had numerous discussions, communications, and meetings with GM concerning the Commission s investigation into GM' s distribution of new GM crash parts, the system itself, and ideas to change the system (CX 7301A, Tr. 3462, 3483-4, 3500, 3517 CX n05A C; CX 7303A-8062--3, 8068, 8170, 8177- , see CX 7313M-O, CX 7314A-B; CX 7316A-B; CX 7317; CX 7318; CX 7332A-E; CX 7355B) (CCPF 281) (RRB 281).

293. During many of these discussions, communications, and meetings, GM dealers, directly or through their associations, have repeatedly urged GM not to sell service GM crash parts to other than GM dealers (CX 7301; CX 7314B; CX 7319; CX 7341A-B; CX 7352A- B; Tr. 3361, 4687-89; see Stipulation Tr. 3376) (CCPF 282). 294. For over a year prior to GM's July, 1975, proposal to the Commission, NADA representatives " debated the crash parts issue with the Federal Trade Commission and General Motors" and . . . repeatedly urged General Motors not to open their warehouses not to sell non-dealers (CX 7314B) (RPF 224). 295. In May, 1975, NADA officials, at a meeting attended by Mr. Estes, GM's president, and other GM officials, indicated their opposition to the opening of GM's warehouses to either IBSs or IWs (Tr. 4687-90; CX 7314) (RPF 225).

296. Mr. Estes, expressed sympathy with NADA's viewpoint namely, ". . . that the present system served the consumer properly and that the dealers had made an investment " but also indicated that GM was under considerable pressure and that more was at stake than merely parts distribution (Tr. 4690) (RPF 226). On June , 1975, Mr. Pohanka, a GM dealer in the Washington, D.C. area who at that time was vice-president of NADA urged GM, by telegram to Mr. Estes, not to open its warehouses (Tr. 4668; CX 7314B) (RPF 228).

::97. A GM representative called the Executive Director of NADA to inform him that GM was about to extend the July 11, 1975 settlement proposal to the FTC (Tr. 3539) (CCPF 289). 298. Immediately after GM's offer was announced, NADA organized to fight it (Tr. 3564). NADA sent a circular to all NADA members which stated that the GM proposal was a serious threat to , 464 Initial Decision dealers and that an increase in wholesale compensation was what was needed (CX 73l4A-C). This was followed by a Mailgram to all NADA members urging them to write Mr. Estes, and ask him to withdraw GM's offer (CX 7349B). In response to this request, a large number of Jetters were sent to GM by its dealers urging withdrawal of the offer (Tr. 4721) (CCPF 290).

299. The July 11, 1975, settlement proposal, among other (58) things, would have led to the direct sale by GM of new GM crash parts at dealer net prices to IBSs (CX7010; ALJX 11; Tr. 3559) (RPF 229). Mr. Hancock, NADA's president at the time, Mr. Pohanka, the vice-president, and Mr. McCarthy, the chief administrative officer (Tr. 3422), were informed of GM's proposal on the same day that it was made to the Commission (CX 7314B; Tr. 3559 0; CX 7321) (RPF 230).

300. NADA "strongly" opposed GM's offer to open the warehouses to lBSs and expressed its opposition to individual Federal Trade Commissioners and to members of the FTC staff (Tr. 356O-2). NADA also made public its opposition through a press release issued July 25, 1975 (CX 7301A-B; Tr. 3480) (RPF 232). 301. NADA's efforts to convince GM to abandon its proposal to sell directly to lBSs consisted of "argument". Mr. Pohanka testified We were very distressed when General Motors made the offer to the Federal Trade Commission that they did. We felt it was not in the best interests of the dealer or the consumer, and told General Motors about that" (Tr. 4717-18). GM informed NADA that it did not intend to withdraw its settlement proposal (CX 7305, Tr. 3563-4). 302. On December 17, 1975, representatives of GM and NADA met to discuss the latest developments regarding GM's crash parts stated thatdistribution system. Mr. Mehan, speaking for GM, getting independent distributors into crash parts would result in greater costs to consumers and cause greater dealer problems (CX 7316A; CX 7324A; Tr. 3524). NADA countered with its Four Point Program which did not include opening GM's warehouses to nondealers (CX 7316B; Tr. 3525-29). The question, insofar as one attendee noted, and he was the only one so noting, was What can GM and dealers do together to keep 'independent distributors' out of crash parts area?" Another attendee disputes the accuracy of the note (Tr. 3519-22; CX 7324B) (CCPF 292) (RRB 292). 303. On February 5, 1976, GM and NADA sent separate settlement proposals to the FTC which were described by one NADA offcial as "essentially the same . The GM proposal was very similar to NADA's Four Point Program (CX 7353B; compare ALJX 13A- Supp. to CX 7012 with CX 7327A-- Point 1). Raising wholesale Initial Deision 99 F. T. compensation to 30% had already been adopted by GM prior to the February settlement offer (ALJX 13G, Supp. to CX 7012). 304. As advocated by NADA, GM's proposal did not include the July, 1975, offer to sell crash parts directly to IBSs but did call for wholesale compensation to be paid on crash part sales across franchise lines a Chevrolet dealer sellng a Pontiac part would be eligible to claim wholesale compensation (ALJX 13G, Supp. to CX 7012) (CCPF 293). (59) 305. NADA did not ask GM's opinion of the NADA proposal (Tr. 3570-71, Tr. 4768-9) (RPF 243) and NADA had no advance knowledge of the February 5, GM proposal (Tr. 3570-72; Tr. 4768). (See Finding 52, 303-).

DISCUSSION The Relevant Markets The Prduct Market 306. Identification of the relevant product market or submarket is the first step in a monopolization case. Brown Shoe Co. v. United States, 370 U.S. 294, 324 (1962). Commission counsel contends that the relevant product market consists of service (new) GM crash part (CPF 53-59). GM objects to isolation of crash part for separate analysis from the rest of the "transportation package" it sells in competition with other manufacturers, on the ground that doing so ignores the often exercised owner option to repair part which have been crash damaged. GM argues that if crash parts are to be so isolated then used crash parts must be included within the relevant market. (RB 5) 307. Counsel to GM also argues that in United States v. Alumi. num Co. of America (Alcoa), 148 F.2d 416, 424-25 (2d Cir. 1945), Judge Learned Hand reasoned that the company took into account that part of its current production would be salvaged in determining what its output of new aluminum should be. GM, the argument goes is merely interested in increasing the sale of new cars and the expected supply of used crash part is not a factor taken into account in the production of new cars (RB 9). But Judge Hand's reasoning that the secondary material market for aluminum curbs prices of new aluminum does not mean that in this case both new and used crash part must be combined in defining the relevant product market. Even if GM does not take the supply of used crash part into account in determining what its output of new cars wil be, that fact does not obviate existence of a separate and distinct new GM crash . . . . . . . 464 Initial Deision part market. The precedents reflect that making that determination is accomplished by examining the product involved and not necessarily by considering how the producers view them. 308. "Cross-elasticity of demand" was the criterion used to identify the relevant product market in United States v. E.I. du Pont de Nemours Co. the "cellophane" monopolization case. 351 U.s. 377 (1955). It was stated that " commodities reasonably interchangeable by consumers for the same purposes make up that ' part of the trade or commerce, monopolization of which may be ilegal" 351 U.S. at 395. In other words, under the cross-elasticity of demand test, if purchasers can substitute the products of one supplier for the (60)products of other suppliers, the products which may be substituted are included in the market for examination. But that standard proved too restrictive to always be used. 309. Seven years later, in Brown Shoe, supra after citing Dupont/Cellophane the Supreme Court said that while there may be broad product markets whose outer boundaries u are determined by the reasonable interchangeabilty of use or the cross-elasticity of demand between the product itself and substitutes for it. ," there also may be "well defined submarkets" within the broader market. 370 U.s. at 325.

310. The Court added that relevant submarkets could be identified by . . . such practical indicia as industry or public recognition of the submarket as a separate economic entity, the product' peculiar characteristics and uses, unique production facilities, distinct customers, distinct prices, sensitivity to price changes and specialized vendors." 370 U.S. at 325. The result was that in Brown Shoe the markets for men, women, and children s shoes were examined because they were economically significant submarkets within the shoe industry.

311. A few years later in General Foods Corp. v. F. T.C. 386 F. 936, 940 (3rd Cir. 1967), cert. denied 391 U.S. 919 (1968) (which was cited by the Commission in Borden, Inc. (Realemon) 92 F. C. 669 784, n.8 (1978) and is the Commission s most recent opinion in a monopolization/monopoly power case) the court again made it clear that the existence of some cross-elasiticity of demand or as the Commission put it in Borden/Realemon some degree of interchangeability, does not foreclose the existence of submarkets identified by Brown Shoe criteria.

312. There is a detailed discussion of the reasons for the development of the seven criteria test in Reynolds Metals Co. F. T.G. 309 F.2d 223, 226-229 (D.C. Cir. 1962), a merger case involving acquisition of a producer and seller having a substantial , Initial Decision 99 F. (33%) market share of the decorative aluminum foil market by a major manufacturer of aluminum. Also see G. Balfour Co. FT.G. 442 F.2d 1 , 11 (7th Cir. 1971) and RSR Corp. v. FT.G., 1979- Trade Cases 450, p. 76 663-4 (13 C. 1). 313. In United States v. Aluminum Co. of America (Alcoa-Rome), 377 U.S. 271 (1964), separate aluminum and copper submarkets were found to exist in the wire and cable industry. Existence of a separate paper insulated power cable submarket within a stipulated insulated wire and cable line of commerce (market) was found in United States v. Kennecott Copper Corp. (Kennecott), 231 F.Supp. 95, 98-100 (S. Y. 1964) aff'd per curiam. 381 U. S. 414 (1965). In United States v. Bethlehem Steel Corp. 168 F. Supp. 576, 593-95 (S. 1958), the iron and steel industry was found to be the relevant broad (61Jline of commerce, but ten specific products (e. hot rolled sheet track spikes, electricweld pipe, oil field equipment supplies) were held to be identifiable submarkets as well. 314. In United States v. Grinnell Corp., 384 U.S. 563, 572-573 (1966) (a leading monopolization/monopoly power case), the Supreme Court said . . . in 2 cases under the Sherman Act, as in 7 cases under the Clayton Act (citing Brown Shoe J there may be submarkets that are separate economic entities. We see no reason to differentiate between 'line' of commerce in the context of the Clayton Act and 'part' of commerce for purposes of the Sherman Act (citing United States v. First National Bank Trust Co. 376 U.s. 665 , 667-68 (1964), a, Sherman Act case)." Also see Columbia Broadcasting System v. FT.G. 414 F.2d 974, 978-79 (7th Cir. 1969), cert. denied 397 U.s. 907 (1970).

315. The consistent thread running throagh the decisions is that the objective and need is to delineate markets which conform to areas of effective competition and to the realities of competitive practice, regardless of which test is used. Balfour, supra 442 F.2d at 11. The approaches to identify broad markets and the submarkets contained with them, are described in Borden Inc./Realemon, supra 92 F. C. at 783--8.

316. By reference to Brown Shoe indicia and the cross-elasticity of demand test in Dupont/Cellophane it is possible to combine crash parts, both new and used, together with the repair of crash damaged portions of a vehicle in a three component, broad relevant market, as counsel to GM suggests. Examining only such a market would be appropriate in a case in which the focus is solely on the alternative ways in which a crash damaged vehicle might be repaired but that is not our focus here. Our role is to determine whether there is a substantial anticompetitive effect on any product market affected g., 464 Initial Decision the acts or practices alleged to be ilegal the distribution by GM of new crash parts.

317. Commission counsel relies on five of the Brown Shoe criteria to separate new and used crash parts into two submarkets: (1) specialized vendors; (2) peculiar characteristics and uses; (3) industry and public recognition; (4) distinct prices; and (5) sensitivity to price changes (CCPF 379-396).

(1) Specialized Vendors 318. With the exception of a very small number of crash parts GM is the sole distributor of new GM crash parts and distributes them to its franchised dealers. GM does not sell used GM crash parts (Findings 12(8)(9), 68). These are obtained only from specialized outlets salvage yards and bumper rechromers (Finding 214). (62) (2) Peculiar Characteristics and Uses 319. The limited availabilty of used (salvage) crash parts for vehicles less than two years old (Finding 235) and the fact that certain service parts exterior moldings, grilles, fenders, bumpers, quarter panels, etc., are seldom available in salvage form (Finding 222) distinguish new from salvage GM crash parts. In addition, dealers sell new crash parts as individual items, whereas salvage yards most frequently sell their product as part of large assemblies (Finding 223) requiring different types of labor for installation.

320. As pointed out by Commission counsel (CCPF 387), distinctions in quality have been held to justify treating two products as being in separate markets. United States v. Pennzoil 252 F. Supp. 962 972-76 (W. D. Pa. 1965). A.G. Spaulding Bros., Inc. v. T.C. 301 F.2d 585, 599 03 (3rd Cir. 1962). The fact that used crash parts are often bent, rusted, irregular, and more diffcult to repair (Finding 225) is a peculiar characteristic which justifies new and used crash parts being considered as separate submarkets. (3) Industry and Public Recognition 321. Individuals in salvage crash parts and bumper rechroming businesses recognize that these industries are separate from the distribution of new crash parts (Finding 225). Separate trade associations exist (Findings 14 and 15) which is further evidence that the two industries are distinct. Bethlehem Steel, supra, 168 F. Supp. at 594; United States v. Citizen Publishing Co. 280 F. Supp. 978, 985 (D. Ariz. 1968), aff'd 394 U.S. 131 (1969). In addition, vehicle owners Initial Decision 99 F. and body shops prefer the use of new crash part (Finding 224), recognizing that quality distinctions may exist, despite possible insurance company preferences for the use of salvage parts (Findings 218, 230, 240).

(4) Distinct Prices 322. There is generally at least a 25% price differential between wholesale salvage crash part and their new counterparts (CCPF 96). Similar price differences have been considered Ustrong evidence" of separate markets. Borden/Realemon, supra, 92 F. C. at 763, citing Brown Shoe; Alcoa-Rome, supra 377 U.S. at 276; Reynolds Metals supra 309 F. 2d at 229; Litton Industries, Inc. 82 F. C. 793, 997 (1973).

(5) Price Sensitivity 323. There is a lack of mutual price sensitivity between new and used GM crash part (CPF 41, 94, 94A, 95). GM uses a list price for part. It is distributed nationwide and does not (63Jtake into account variations in prices for salvage crash parts, marketed on a local level, for which the prices change frequently (CPF 95). 324. Counsel to GM contends that use of Brown Shoe standards wil result in an even broader definition of the product market than Commission counsel advocates in that it would include used GM crash part (RB 8). The reasoning of counsel to GM is not persuasive. It is true that insurance companies, salvage operators, and installers understand that new and used crash parts are both options to be considered in the repair of crash damaged cars (RPF 191-202), but there is stil recognition that there are two separate and distinct systems of distribution.

325. Salvage part prices and the cost to repair rather than replace are taken into account by insurers in arriving at the figure they wil pay to have a vehicle fixed. The prices for salvage part or for repairs, if there is any connection, normally " follow" rather than lead" the prices GM establishes for new crash parts but model year of the car also affects the "used" price (Findings 227-8, 236). The lack of mutual price sensitivity (i. one product is price sensitive to another but not vice versa) has been held to be evidence of separate markets. In Dean Foods Co. 70 F. C. 1146 (1966), the Commission found that the price of retail milk moved as the price of wholesale milk moved. However, wholesale milk prices were not sensitive to retail milk prices. In finding separate markets the Commission stated:

What is of significant determinative value in determining the proper scope of a . . ./ . . 464 Initial Decision market involving the same product is whether the price sensitivity which does exist is mutual, whether it is generate equally by both sectors or whether, on the other hand the competitive forces are all generated primarily in one sector. 70 F. C. at 1258 (CCPF 395). This lack of mutuality is also evidenced by the fact that salvage crash parts are 75% to 90% of list in the first three model years and decline to 25% to 50% of list thereafter (Tr. 1747-48).

326. Although both types of crash parts are used to repair damaged cars, this does not negate the fact that two separate markets might be found for monopolization purposes. For example the Commission adopted the following language in BordenlReaLemon, supra, 92 F. C. at 762, 832. (R)eognizing that fresh lemons and processed lemon juice are used for many of the same purposes by the public, does not dictate that they must be placed in the same product market where serious, important and (64)economically substantial distinguishing characteristics differentiate the products. In that case, fresh lemons and processed/bottled lemon juice were found to be in different product markets because the bottled product had limited use due to its distinctive taste and the additives it contained, whereas fresh lemons were less convenient to use, subject to spoilage, and had a higher cost. This conclusion was reached without resort to the Brown Shoe criteria. 92 F. C. at 788. The differences between new and used crash part are equally significant.

327. Counsel to GM suggests that "reasonable interchangeabilicross-elasticity" should be the test used to identify the product market here. I do not agree. Two products may have reasonable interchangeability of use or cross-elasticity of demand, but welldefined submarkets stil may exist within a broad market, and they may be product markets for antitrust purposes. BordenlReaLemon supra 92 F. C. at 762, 832.

328. It has often been held that new products may be separated from their used or recycled counterparts in determining the relevant product market. Alcoa, discussed at p. 59. Also see Avnet, Inc. F T. C., 511 F.2d 70 (7th Cir. 1975); United States v. Paramount Pictures 334 U.S. 131 (1948); RSR Corp., supra. 329. Lastly, the fact that not all of the Brown Shoe criteria have been used in defining the relevant market is not significant. There are a number of precedents (e.g., Alcoa-Rome, Kennecott) for the proposition that not all, or even most, of them need be taken into account. United States v. Continental Can Co. 378 U.S. 441, 456-7 (1964); General Foods Corp., supra 386 F.2d at 941; Columbia Initial Decision 99 F. Broadcasting System, supra 414 F.2d at 979. The new GM crash parts market is "suffciently inclusive to be meaningful in terms of trade realities. Crown Zellerbach Corporation v. F T.G. 296 F. 800 811 (9th Cir. 1961), cert. denied 370 U.S. 937 (1962). Consequently, in this case, new GM crash parts comprise the relevant product market.

The Geographic Market 330. The geographic market which one must examine in order to determine whether the monopolization alleged is illegal may be identified in much the same way as the product market. In Brown Shoe the Supreme Court said that the ". . . criteria to be used in determining the appropriate geographic market are essentially similar to those used to determine the relevant product market'" The geographical market selected must'" both correspond to the commercial realities of the industry and be economically significant ... (footnote omitted). (AJlthough (65Jthe geographic market in some instances may encompass the entire Nation, in some other circumstances, it may be as small as a single metropolitan area." 370 U. at 336-37.

331. What is very clear from the precedents is that the geographic market to be examined need not be marked off in metes and bounds. United States v. Pabst Brewing Co. 384 U.s. 546, 549 (1966); du Pont/Cellophane 351 U.S. at 395.

332. The geographical effects of alleged violations of the antitrust laws have been considered by the Supreme Court and lower courts with reference to both broad geographic markets and submarkets within the broad area, in basically the same manner as in the case of product markets. United States v. Kimberly-Clark Corp. 264 F.Supp. 439, 455-56 (N. Cal. 1967); Bethlehem Steel Corp., supra, 168 Supp. at 601--2.

333. In Grinnell, supra 384 U.s. at 576, the court said " . . . the relevant market for determining whether the defendants have monopoly power is not the several local areas which the individual stations serve, but the broader national market that reflects the reality of the way in which they built and conduct their business. 334. Even if GM did not actually sell its crash parts in every state, which is contrary to the evidence in this case, there are numerous precedents to the effect that a national market may exist. See FT.G. v. Procter and Gamble Co. 386 U.S. 568, 571-72 (1967); Pabst, supra 384 U. S. at 549-551; A. G. Spaulding, supra 301 F.2d at ),/ 464 Initial Decision 607; Kimberly-Clark, supra, 264 F. Supp. at 454-58; British Oxygen Company Limited, et al., 86 F. C. 1240, 1346-7 (1975). 335. The nation as a whole most assuredly is significant economically and is the area where the effect of the monopoly/monopoly power on competition is direct and immediate. The Supreme Court has held this to be an appropriate "section of the coungeographic market" insofar as antitrust violations are concerned. Philadelphia NationalBank, 374 U.S. 321 357-362 (1963). The Commission did the same in Borden/Realemon, supra, 92 F. at 675, 832.

336. In this case the United States as a whole is the relevant geographic market. This is because new GM crash parts are marketed nationally (Finding 67). This fact alone warrants considering the nation as the relevant geographic market. See Commission Opinion in Beatrice Foods Co., 86 F. C. 1 60 (1975). Antitrust Case Decisions as Precedents for Section 5 FTA Decisions 337. With regard to the fact that many of the cases cited as precedents have involved charges of several different (66)violations of the antitrust laws or the Federal Trade Commission Act, the standard for outlining the relevant markets is no different whether a case has been brought under the Sherman, Clayton or Federal Trade Commission Acts. Borden/Realemon, supra, 92 F. C. 784, citing Luria Bros. Co. 62 F. C. 243 , 604 (1963), aff'd 389 F.2d 847 (3rd Cir. cert. denied 393 U.S. 829 (1968); Columbia Broadcasting, supra, 414 F.2d at 978-79.

Monopolization/Monopoly Power 338. Monopoly power is the power to control prices or to exclude competition. American Tobacco Co. v. United States, 328 U.S. 781 811 (1946). It usually can be inferred from possession of a predominant share of the relevant market. Grinnell, supra 384 U. S. at 571; Dupont/Cellophane, supra, 351 U.S. at 391. Having identified the relevant product and geopraphic markets, the respondent's share of those markets is considered to determine whether the firm s actions constitute monopolization or whether the firm has monopoly power. Grinnell, supra 384 U.S. at 571.

339. !!Size is of course an earmark of monopoly power United States v. Griffith 334 U.S. 100, 107 n.lO (1948). GM's size (Finding 65) and the fact that its crash parts are made for GM cars only , Initial Decision 99 F. T. (Finding 114) make it clear that GM has monopoly power, insofar as new GM crash parts are concerned.

340. The market share of the relevant product(s) in the relevant geographic market, that is the annual sales figure, is used in making this determination. It has been held that the share probably must be greater than sixty percent in order for monopoly power to exist. Alcoa, supra 148 F.2d at 416. Alcoa had 90% of the relevant market in virgin aluminum. 148 F.2d at 424. Here, GM approaches a 100% share because it supplies practically all of the new GM crash part which are distributed (Finding 68).

341. GM is able to (and does) set the prices of the crash part it sells without any real concern for the near term reactions which dealers or others might have to the increase in price. No doubt GM has concerns about possible long term effects price changes may have on car buyers' decisions. There was testimony that competitors prices for crash part are noted when price increases are recommended to higher level GM employees who make the decisions on prices (Finding 199). But there was no persuasive evidence to the effect that the adverse reactions of dealers, other repairers of amaged autos, or consumers are either considered or have a controlling influence on whether an increase should occur when price increases are being decided upon.

342. Thus, GM has a monopoly and monopoly power by virtue of the fact that, except for a very few part, it is the (67)exclusive source of supply for new GM crash parts. This provides GM with a dominant market share and a position well insulated, even isolated from competition for sales of new GM crash parts. Were/Are GM' s Practices an Abuse of Monopoly Power Illegally Monopolistic? 343. A finding of a monopoly in the relevant market is not enough to constitute a violation of antitrust law of the Sherman Act (15 D. C. 2) because having a monopoly or monopoly power as such is not illegal per se. It is monopolization, attempts to monopolize, and abuses of monopoly power which are prohibited. "According to the court interpretations of this (Sherman Act) statutory language that have been handed down since 1890, Congress' failure to outlaw monopoly as an end result and focus instead on the means by which a monopolist was thought to get there was no accident or oversight: It didn want to make monopoly ilegal; it simply wanted to make sure that, wherever monopoly does appear, it wil have been acquired fairly. (Emhasis added)". Mueller Entrepreneurial Education, 2 , 464 Initial Deision Antitrust L. and Econ. Rev. 4 (I979). "To monopolize is not simply to possess a monopoly: the word implies some positive drive, apart from sheer competitive skils, its size and power in the market." Neale The Antitrust Laws of the USA (2nd Edition, 1970) 92, 93. 344. "The Sherman Act was intended to secure equality of opportunity and to protect the public against evils commonly evident to monopolies and those abnormal contracts and combinations which tend directly to suppress the conflct for advantage called competition-the plan of the contending forces ordinarily engendered by an honest desire for gain. The statute did not forbid or restrain the power to make normal and useful contracts to further trade by resorting to all normal methods, whether by agreement, or otherwise, to accomplish such purpose. United States v. American Linseed Oil Co. 262 U.S. 371, 388 (1923) (citations omitted). 345. The acts and practices charged need not in themselves be independently unlawful or predatory to constitute violations of Section 2, but merely being a large company with a monopoly share of the relevant product market is not suffcient. United States Griffith 334 U.S. 100, 105 (1948); Alcoa, supra 148 F.2d at 431-32. However, as the Commission said in Borden/ReaLmon, supra One point made clear by the Alcoa case is that the conduct of firms with monopoly power is viewed differently from that of firms without such power (Emphasis added)." 92 F. C. at 794. 346. "The existence of monopoly power- monopoly in the concrete' Standard Oil Co. of New Jersey v. United States (221 U.S. 62 (1911)) does not by itself prove the offense of (68)monopolization.. . . the offense is the existence of the power to raise prices or exclude competition coupled with the purpose or intent to exercise that power.' The requisite intent for this purpose is not a ' specific' intent to monopolize, but rather a conclusion based on how the monopoly power was acquired, maintained or used." Report Of The Attorney General's National Committee to Study the Antitrust Laws March , 1955, p. 55, citing Griffith, supra 334 U.S. at 107; American Tobacco, supra 328 U.s. at 809.

347. Consequently, for a violation of the antitrust laws to be found when a firm has a monopolist's share of the relevant markets or when the firm has monopoly power, the firm must be shown to have conducted its business in such a way that it went beyond the very fine line dividing acceptance of its position and taking actions which would unduly stifle, frustrate or foreclose competition. For example, Alcoa, by expanding existing plants and constructing new facilties with the goals of building ahead of demand and unduly taking the initiative in dealings with purchasers, suppliers, and . . .

Initial Decision 99 F. competitors so as to preempt any diminution of its power was found to be an ilegal monopolist under the Sherman Act. Alcoa, supra, 148 2d at 431-32. Also see United Shoe Machinery, 110 F. Supp. 295 341 (D. Mass 1953); Grinnell, supra 384 U.S. at 576. 348. Insofar as the Federal Trade Commission Act is concerned neither an offensive practice nor conduct constituting a violation of the Sherman or Clayton Acts need exist for a violation to be found. The reason is that the Commission is not bound to follow antitrust standards as strictly as the courts must in cases brought under the Sherman and Clayton Acts. Fashion Originators ' Guild of America Inc. v. F. T. G., 312 U.S. 457, 466-7 (1941); Sperry Hutchinson Co. F T. 432 F.2d 146, 150 (5th Cir. 1970) rev d on other grounds, 405 U.s. 233 (1972).

349. Controllng Commission and court precedents tell us that monopolization is illegal: if (1) the possessor has monopoly power in the relevant market and (2) that power was acquired or is maintained wilfully "as distinguished from growth or development as a consequence of a superior product, business acumen, or historic accident. Borden/Realemon, supra, 92 F. C. at 788, citing Grinnell, supra 384 U.S. at 570-71.

350. In addition to demonstrating that monopoly power exists, it must be shown that the power was wilfully acquired or maintained. Thus, in Borden/Realemon the wilful nature of Borden s actions was shown by evidence that the firm took steps to ensure that its monopoly position would not be lost or eroded, and engaged in acts and practices designed to frustrate competition." 92 F. C. at 792. These steps included: (1) spurious product differentiation that enabled it to command a substantial price premium (p. 793); (2) manipulation of this (69Jprice differential (p. 793); (3) sacrificing somewhat higher prices over the short-run to assure continued monopoly returns over the long haul (p. 795); (4) the use of geographically discriminatory promotional allowances (p. 795-96); (5) demanding a price considerably in excess of that of competing brands in some areas and reducing prices selectively in areas where it wished to suppress emerging competition (p. 797); and (6) sellng to competing customers in the same local market at different prices without cost justification (p. 798-99).

351. To the same result see Grinnell where the Court noted that the company: (1) used restrictive agreements to preempt its competitors; (2) used pricing practices to contain its competitors; and (3) acquired competitors in achieving its monopoly position unlawfully. 384 U.S. at 576.

352. No practices similar to those in which Borden and Grinnell . . . . . .

464 Initial Decision engaged have been shown to have been used by GM in order to acquire or maintain its monopoly in new GM crash parts. GM does require suppliers using its dies and inserts to sell their output only to GM (Finding 68), but the evidence does not show that GM takes action to inhibit the production of dies and inserts by others who might wish to make and sell crash parts or that GM seeks to enforce any design patents or common law rights it possesses in that regard. For example, a witness testified that his firm makes new crash parts for GM's Corvette without interference (Tr. 1895). 353. Contrary to the position of Commission counsel the facts here are very different from those found in United States v. General Motors Corporation 121 F. 2d 376 (7th Cir. 1941) and in United States v. General Motors Corp. 384 U.S. 127 (1966). Counsel to ASC makes the same argument about the latter case (lab, pp. 12-13). In the earlier case GM sought to extend its monopoly into another field (i. the wholesale and retail financing of GM autos) by conspiring with General Motors Sales Corporation, General Motors Acceptance Corporation, and General Motors Acceptance Corporation of India- , Inc. and by coercing dealers to deal with the finance companies. 121 F.2d at 399. In the latter case GM combined with some of its dealers to stop other dealers from sellng cars to discount houses, and policed their activities to insure that they did so. Justice Fortas found H . . a classic conspiracy in restraint of trade: joint collaborative action by dealers, the appellee associations, and General Motors to eliminate a class of competitors by terminating business dealings between them and a minority of Chevrolet dealers and to deprive franchised dealers of their freedom to deal through discounters if they so choose." 384 U.s. at 140. No such evidence was disclosed in this case.

354. GM does have an inevitable monopoly in new GM crash part. As to these some are natural, and others are 'thrust upon' their owners." But, as the Commission cautioned (70Jin Borden/Realemon:

where a firm with monopoly power interferes with natural economic forces which would otherwise dissipate its monopoly, the law rightfully condemns it. 92 F. C. at 795.

355. The evidence here does not show that GM has interfered with natural economic forces which would otherwise dissipate its monopoly. Thus, insofar as the charges of monopolization/monopoly power over new GM crash parts are concerned GM ". . . falls within the exception established in favor of those (monopolists) who do not . . .

Initial Decision 99 F. seek but cannot avoid, the control of a market." Judge Learned Hand in Alcoa, supra 148 F.2d at 431.

356. The teaching of Alcoa is that monopolization may be ilegal not because a firm is progressive, but rather because it acted with calculation to head off every attempted entry into the field. The case is not to be interpreted as penalizing enterprise, instead it declares ilegal those monopolies which are maintained by policies intended to discourage, impede or even prevent the rise of new competitors. Attorney General's Committee Report, supra at 60. The evidence here does not show that GM has discouraged, impeded or prevented the rise of new competitors in the new GM crash parts market. The concurring opinion of Commissioner Pitofsky in Borden/Realemon supra details the law with regard to the maintenance of monopoly power by use of unreasonably exclusionary behavior. 92 F. C. at 820-21.

357. I am convinced that GM's restricting suppliers using its dies to sales to GM (Finding 68) is not unreasonable. There is no other showing in this case which persuasively evidences behavior by GM to exclude other manufacturers from the relevant market. GM's Right to Choose Its System for Distributing New Service Crash Parts 358. Commission counsel contends that by sellng new crash parts only to its franchised dealers, GM is using its monopoly power in violation of the Sherman Act, (or in violation of its spirit, and therefore ilegally under Section 5 FTCA) because GM forestalls competition by its refusal to deal with other than its franchisees (CCPF 419). The following language from United States v. Arnold Schwinn Co. is quoted to support his proposition that a manufacturer must deal with all those who seek to have it do so, when products competitive to the manufacturer s are unavailable. (A) manufacturer of a product other and equivalent brands of which are readily available in the (71)market may select his customers, and for this purpose he may franchise" certin dealers to whom alone, he wil sell his goo. (Citations omitte). If the restraint stops at that point-if nothing more is involved than vertical confinement" of the manufacturers' own sales of the merchandise to selected dealers and if competitive products are readily available to others, the restriction, on these facts alone would not violate the Sherman Act. 388 U.S. 365, 376 (1966).

359. Taken by itself, this language appears to confer an obligation to deal upon a manufacturer, but when examined in context it is clear that what the Supreme Court was talking about was part of the , 464 Initial Decision broad outline of conduct permissible under the Sherman Act. The statement was meant merely to identify one end of a spectrum, with price fixing, conduct which never is permissible, identifying the other end.

supra, holds that a manufactur- 360. I do not agree that Schwinn, er of a product, having a monopoly in the relevant product market Schwinn, themust do business with all wiling customers. In Supreme Court merely held that the manufacturer imposed vertical restrictions on the resale of its goods by franchisees which constituted ilegal restraints of trade. Schwinn s conduct was not a mere refusal to trade with those who were not franchisees. It imposed territorial limitations on resales by distributors and confined resales by franchisees and distributors of bicycles to franchised retailers. 365 U.s. at 370-71. No such limitations have been shown in this case. In fact, GM has encouraged dealers to sell to IBSs by having the wholesale compensation plan apply to such sales. 361. I agree with counsel to GM when he points out that Any analysis of the GM system must begin with the recognition that a manufacturer, absent any purpose to monopolize may choose the (R. Br. 18). United Statescustomers with whom it wil deal." v. Whirlpool Corp., 579Colgate 250 U.S. 300, 307 (1919). Oreck Corp. 2d 126 (2nd Cir.), cert. denied 58 L.Ed. 2d 338 (1978); Schwing Motor Co. v. Hudson Sales Corp. 138 F. Supp. 899 (D. Md. aff'd, per curiam 239 F. 2d 176 (4th Cir. 1956), cert. denied 355 U.S. 823 (1957); Quality Mercury, Inc. v. Ford Motor, Co. 542 F.2d 466 (8th Cir. 1976), cert. denied 433 U.S. 914 (1977).

362. Further analysis of the system GM uses for distributing new GM crash parts start from the premise that, absent a purpose to monopolize or an effect producing an unreasonable restraint on trade, GM may choose its system of distribution. In other words, it is stil the law that a supplier may choose the customers with whom it will do business. (72) 363. The rationale for the principle as it applies to this case is set forth in Schwing, 138 F. Supp. at 902. Every manufacturer has a natural and complete monopoly of his particular product especially when sold under his own private brand or trade name. Arthur v. Kraft- Phenix Cheese Corp. 26 F.Supp, 824, 828 (D.C. Md. (1937)). If he is engaged in a private business, he is free to exploit his monopoly by sellng his product directly to the ultimate consumer or through one or more distributors or dealers, as he may deem most profitable to him. If he chooses the latter method, he may exercise his own independent discretion as to the parties with whom he will deal. This is a common law right which the antitrust laws have not destroyed. (citing Colgate, supra; F.T.G. Raymond Bro. Clark Co., 263 U.S. 565 (1924); Times-Picayune, supra; Great Atlantic & Pacific Tea Co. v. Cream of Wheat CD. 227 F. 46, 49 (2d Cir. (1915))). A refusal to deal g.

Initial Decision 99 F. becomes illegal only when it produces an unreasonable restraint of trade or monopoly forbidden by the antitrust laws. Kiefer-Stewart Co. v. Joseph E. Seagram & Sons, Inc. 340 U.S. 211 (1951); Lorain Journal Co. v. United States, 342 U.S. 143 (1951).

364. Thus, in the absence of any purpose to create or maintain a monopoly, or effect constituting an unreasonable restraint, it is legal for a manufacturer, engaged in an entirely private business, to exercise his own independent choice of the parties with whom he wil deal. Colgate, supra, 250 U.S. at 307. But this right to choose the customers is neither absolute nor exempt from oversight by the government. And, if the right is exercised with the purpose or effect of monopolizing interstate commerce, or otherwise is unreasonable such exercise is unlawful. Lorain Journal Co. , supra, 342 U. S. at 155. 365. An individual refusal to deal, or, as here, GM's refusal to sell crash parts to anyone other than a GM dealer, is a circumscribable general right. If the refusal were accompanied by predatory conduct or agreement with other manufacturers, or if the general right were exercised to impede competition on the basis of price or for a monopolistic purpose for market control, the refusal would be if illegal. Times-Picayune, supra 345 U.S at 622-23. For example, GM refused to sell directly to IBSs or IWs in order to influence prices or to maintain or to extend power over the market that would illegal. Banana Distributors, Inc. v. United Fruit Co. 162 F. Supp. 32 37 (D. Y. 1958). However, there is no evidence of any such conduct or effect in this record.

366. In the same year that Alcoa was decided, a District Court held in an often cited case that a monopolist had violated (73jthe law in that its acts were not directed to any legitimate business venture , in fact, aoriented toward furthering its own business, but were calculated attempt to monopolize government contracts for a certain product. A refusal to deal, while it may be lawful per se cannot be used in order to achieve an ilegal result. United States v. Klearfax Linen Looms 63 F.Supp. 32, 39 (D.C. Minn. 1945). Also see Eastman Co., 273 U.S. 359, 375 (1927), Kodak Co. v. Southern Photo Materials where Kodak ilegally sought to move into retailing through use of its monopoly in film by refusing to continue to sell photographic supplies to a retailer at dealers' discounts. 367. The question to be answered, in the absence of predatory motives and with the presence of legitimate business purposes (such as a better way to distribute new GM crash parts in order to promote the sale of new cars, or to stabilize dealer outlets or to augment profits-GM' s motives here), simply is whether the effect of GM' refusal to deal with other than its dealers is substantially adverse to . . .

464 Initial Decision competition. Schwinn, supra 388 U.s. at 375. The evidence shows that the answer to the question is that the refusal to deal, in itself, is not substantially adverse to competition. 368. Neither Schwinn nor Continental TV, Inc. v. G. T.E. Sylvania 433 U.S. 36 (1977) (which reversed Schwinn in part), imposes upon GM any duty, absent an intent to monopolize, to deal with all wiling customers. These cases are not the only relatively recent expressions of that concept. For example, in Venzie Corp. v. United States Mineral Products Co., 521 F.2d 1309 (3rd Cir. 1975), the court upheld a refusal to sell to plaintiff the only fireproofing material that would be approved by the city of Philadelphia, which refusal caused the plaintiff to lose a building contract. The court said To adopt plaintiffs' position would revolutioniz the antitrust field. Every refusal by a franchisor to deal with one not a franchise would automatically lead to a per se violation of the Sherman Act if the franchisor s product possessed the "desirability to consumers" or "uniqueness" which have ben found suffcient to establish the necessary economic power of the tying product. It would, moreover, amount to a substantial undercutting of the Col ate doctrine validating unilateral refusals to deal in the absence of a monopolistic purpose. Neither precedent nor policy suggests that such a reordering of the antitrust implications of busines."i behavior is in order- 521 2d at 1318.

369. An important factor in Venzie was that defendant company made no effort to use the economic power it possessed as the sole manufacturer of the fireproofing material to enter the business of fireproofing application. 521 F.2d at 1317. This (74Jis also true of Ford' s actions in not sellng crash parts to a wholesaler, as described in FLM Collision Parts v. Ford Motor Co. 543 F.2d 1019 (1976). It also is true of GM's actions as shown here. 370. The desirability of not impinging unduly on the freedom to choose customers was expressed by Judge Hand in Alcoa, supra, 148 2d at 427-28. ". . . (IJt is of the essence of competition that the manufacturer or wholesaler should and does have wide freedom in maintaining the quality of his distribution system." This concept is found in Colgate and in Judge Hand' Alcoa decision. Alcoa recognized as ". . . the most eloquent statement of the law of monopolization. BordenlReaLemon, supra 92 F. C. at 793. 371. One might argue, as Commission counsel seems to (CCPF 419), that there are no "other and equivalent brands" of new service GM crash parts available. It sounds plausible due to the fact that such parts are designed for use on GM vehicles only and are almost entirely manufactured by or for GM. Hence, it is argued that ilegal monopolization exists in the refusal to deal directly with IBSs and IWs. But that argument ignores the "natural monopoly" each ), Initial Deision 99 F.

manufacturer has over its products, which standing alone is not ilegal. Further, if followed to its ultimate conclusion, the argument would lead to an assertion that a manufacturer is an ilegal monopolist whenever, though free to do so, no one else produces part to repair the manufacturer s product, regardless of whether it is an auto, camera, refrigerator, radio, TV set. watch or what-haveyou. Neither antitrust nor trade regulation law goes that far. 372. There are a number of cases holding that a refusal to deal violates the Sherman Act when the monopolist was or would be in competition with the aggrieved party at some point in the distributive chain. See Gamco, Inc. v. Providence Fruit and Product Building, Inc. 194 F.2d 484 (1st Cir. cert. denied 344 U.S. 817 (1952); United States v. Otter Tail Power Company, 410 U.S. 366 (1973); United States v. Terminal Railroad Association of St. Louis 224 U.s. 383 (1912); Associated Press v. United States 326 U.S. 1 (1945); Kodak, supra; Klearfax, supra. But GM is not in competition with the aggrieved parties here.

373. Complaint counsel contends that GM has a meaningful presence at the wholesaling level through the temporary interest it retains in new dealership franchises financed through its Motors Holding Divison (MHD) (CPF 441). I do not agree. This presence is minimis. The franchisee runs the dealership and the interest GM has is limited in duration to an average of five to six years (RX 34) (RPF 187). Further, only a very small minority of GM dealers are MHD dealers and their number is decreasing. Between 1970 and 1977, the number of such dealerships decreased from 449 to 345 (RX 34). In 1977, the 345 (75)constituted only 3% of the 11 660 GM car dealerships (RX 34; RX 40). MHD dealerships only accounted for an estimated 4.3% of total GM dealer wholesale parts sales and an estimated 5.3% of total GM dealer body shop sales (RX 34; RX 38; RX 40).

374. What the court said of Ford in FLM, supra is true of GM. There is no convincing proof". . . that Ford, as distinguished from its dealers, had any significant share of that market, much less that it was using its monopoly at the manufacturing level to extend such a share." 543 F.2d at 1030.

375. Statements of Commission counsel regarding a symbiotic relationship between GM and its dealers notwithstanding (CPF 441), , any specific dealer, and the dealers as a group, are separate entities. Their actions in furthering their separate and mutual interests, as shown in the record, do not equate to a horizontal boycott, constituting either a per se or rule of reason violation of the Sherman Act or a spirit-or-intent violation of Section 5 of the FTCA. 464 Initial Deision 376. The fact that GM is a very large corporation (Findings 62 and 65) has been considered. But size in itself does not create an unlawful monopoly and successful business operation doesn t either. To hold otherwise would frustrate the principal purpose of the antitrust laws, which is to preserve a system of free competitive economic enterprise and to protect the public against the evils of monopoly and monopoly power which unreasonably suppress or restrain interstate trade or commerce. Kansas City Star Company United States, 240 F.2d 643, 658 (8th Cir. 1957), citing 58 CJS Monopolies Section 18. However, as well as assuring to the consumer the benefits flowing from free competition, another significant purpose of the antitrust laws is to protect the individual businessman. See Judge Bazelon s dissent in Packard Motor Car Co. Webster Motor Car Co. 243 F.2d 418, 422 (D.C. Cir. 1957). 377. Insofar as the system GM uses for distributing crash parts is concerned, no persuasive evidence has been introduced of either a predatory intent or substantially adverse effect on competition attributable to the refusal to sell new GM crash parts to anyone other than GM dealers. The evidence does indicate that GM uses its system to sell crash parts exclusively through its dealers because of their mutual interest in crash parts being readily available (Findings 59 and 94), and that GM does not set or monitor the prices at which crash part are sold (Finding 198).

378. Contrary to Commission counsel's view, I do not believe The v. F. T.Peelers Co., 65 F. C. 794 aff'd in part, sub. nom. La Peyre 366 F.2d 117 (5th Cir. 1966), is precedential. In that case respondent engaged in deliberate and obvious price discrimination between two groups of customers (processors on the west coast and those on the southern coast), (76Jin the leasing of shrimp peeling equipment based on the customers' location, clearly with the goal of protecting its own canning interests at one of the locations. 65 F. C. at 839. The predatory motives found in that case are not present here. (1968), 379. I also believe that FT v. Texaco 393 U.S. 223, Atlantic Refining Co. v. 381 U.S. 357 (1965), and Shell Oil Co. FT 360 F.2d 470 (5th Cir. 1966), the so-alled TBA cases, are notFT dispositive. The Supreme Court and the Fifth Circuit stated that certain marketing arrangements for tires, batteries, and accessories were illegal because the oil companies received commissions from leading tire companies for coercing the oil companies' franchised gasoline stations into buying their stock of those products from the tire companies. The analogy between the practices found in the TBA cases and GM' s wholesale compensation plan to induce dealers to sell to IBSs at dealer net is too tenuous to support a coercion theory here. . . .

Initial Decision 99 F. I am convinced, however, of the illegal nature of the wholesale compensation plan as distinguished from GM's (1) monopoly power over its crash parts as well as (2) the distribution system itself. Does GM's Relationship With Its Dealers Constitute A "Contract Combination or Conspiracy " In Violation of Section of The Sherman Act? 380. Commission counsel contends that GM reached a " mutual understanding and agreement with its dealers, that GM would refuse to sell service GM crash parts to IWs and 1BSs (and that this agreement) violates Section 1 of the Sherman Act and Section 5 of the Federal Trade Commission Act under either a vertical combination analysis, rule of reason analysis, or horizontal combination analysis (citations omitted) (CCPF 398). But, as pointed out by counsel to GM, there is no evidence of any agreement, tacit or formalized, between GM and its dealers that would restrict GM from selling its service crash parts to nondealers, and there is no persuasive evidence in the record indicating that GM' s refusal to do so resulted in any way from dealer pressure (RB 21). 381. In support of the claim of a vertical combination in violation of Section 1 of the Sherman Act Klor s Inc. v. Broadway-Hale Stores Inc. 359 U. S. 207 (1959) and Hershey Chocolate Corp. 28 F. C. 1057 (1939), aff'd 121 F.2d 968 (3d Cir. 1941), are cited by Commission counsel, but they are inapplicable to this case. 382. In Klor's respondents were a chain of department stores and ten national manufacturers and their distributors. The evidence established that a conspiracy existed and that action was taken to keep the manufacturers from sellng to Klor, or to make such sales only at discriminatory prices and with highly unfavorable terms. 359 S. at 208. Justice Black (77)found that Broadway-Hale used its monopolistic" buying power to induce the manufacturers to agree to its plan. 359 U.S. at 209. In addition, he expressly distinguished the plan from a situation where a manufacturer and a dealer agree to an exclusive distributorship. 359 U.s. at 212. No analogous situation f GM bowing to dealer s clout-was shown in this case. To the contrary, GM did offer, in its settlement proposal of July 11 1975, to sell directly to IBSs, despite strong dealer opposition (Findings 299 and 300) (RPF 229-239). Commission counsel's reliance on Hershey, also does not support his position. In that case, Hershey agreed to sell vending machine size bars of its product only to the three largest vending machine companies in order to keep their business. There is no evidence in this record showing that 464 Initial Decision dealers possessed the buying power to induce GM to act involuntarily. Hershey Chocolate 121 F.2d at 970 (RB 21). 383. In sellng service crash parts only through franchised dealers, GM has not shown the intent to limit competition which was present in the situations described in the cases referred to on pages 72 and 73. Nor has GM in any way restricted the freedom of its dealers to do business with any customers with whom they might wish at whatever price they select. On the contrary, the wholesale compensation plan was created to encourage and to induce franchised dealers to sell to IBSs, who are, in essence, the dealers competition (Finding 35).

384. Under the Sherman Act, contracts and combinations which are an unreasonable restraint of trade are prohibited and violations of that act have long been held to be violations of the Federal Trade Commission Act. C. v. Cement Institute, 333 U.S. 683, 691- (1948). Certain practices ". . . because of their pernicious effect on competition and lack of any redeeming virtue are conclusively presumed to be unreasonable and therefore ilegal without elaborate inquiry as to the harm they have caused or the business excuse for their use. Northern Pacific Railway Co. v. United States 356 U.S. 1 5 (1958). Price fixing agreements, divisions of markets, group boycotts, and tying arrangements are practices which are ilegal per se. Northern Pacific, 356 U.S. at 5. However, the actions of GM and its dealers clearly neither rise to a group boycott, nor are they otherwise upernicious . GM's actions have therefore been examined under the "rule of reason whether the negative effects of GM' system of crash parts distribution outweigh any positive benefits to competition. Commission counsel suggests that GM's system provides no benefits to competition (CCPF 406). 385. Vertical restrictions are usually justified on the basis of the stimulation of interbrand competition even as they may work to reduce intrahrand competition. Commission. counsel contends that because GM, for all practical purposes, has a 100% share of the relevant product market, interbrand competition (78Jdoes not exist and therefore cannot be benefitted by the vertical restraints GM has included in its system of distribution (CCPF 403). 386. Counsel to GM responds, and I agree, that the object of the system, to provide optimum serviceability for GM automobiles so that owners wil be favorably disposed to buying another one, is another benefit to be weighed when considering the negative effect of such restrictions (RB 19 20).

387. The Supreme Court has recognized the existence of other redeeming virtues" derived from the use of vertical restraints 546 FEDERAL TRAE COMMISSION DECISIONS Initial Decision 99 F. which enable one manufacturer to compete more effectively against another. Such restraints can be used to induce retailers to engage in promotional activities or to provide service and repair facilities necessry to the effcient marketing of their products. Service and repair are vital for many products, such as automobiles and major household appliances. The availability and quality of such services affect a manufacturer s good will and the competitiveness of his product. Beause of market imperfections such as the so-alled free rider effect these servces might not be provided by retailers in a purely competitive situation, despite the fact that each retailer s benefit would be greater if all provided the services than if none did. (Emphasis added). GTE/Sylvania, supra 433 U.S. at 55.

388. As the Supreme Court stated in GTE/Sylvania a manufacturer generally will prefer a competitive market situation, because this wil: (1) lower retail prices by lessening the margin between the price to dealers and the dealers' resale price; and (2) increase sales and revenues to the manufacturer. GTE/Sylvania, supra 433 U.s. at , n.24.

389. In addition, the fact that GM's system is the one in use by all domestic and foreign competitors (RPF 102-105, 111-119), indicates that the automobile manufacturing industry adopted the system because it believed it to be the most effcient one. 390. I am not persuaded by Commission counsel's argument that , usecure in its position as a monopolist " would sacrifice efficiency in its crash parts distribution system merely to build dealer loyalty and thus increase new car sales. Loyalty to GM is not the major incentive for dealers to maximize new car sales. The financial benefit to dealers who sell additional cars is more than sufficient. In addition, and as noted above, an efficient system of repair of products is a substantial lure for prospective buyers. It is diffcult to (79Jbelieve that GM would conspire with its dealers to the detriment of its own sales of new automobiles. 391. Counsel to GM points out that if any "joint action" can be found between GM and its dealers, it was for the purpose of influencing the government, and thus protected under the Noerr- Pennington doctrine " (RB 22). Eastern Railroad Presidents Conference v. Noerr Motor Freight, Inc. 365 U. S. 127 (1961) and United Mine Workers of America v. Pennington 381 U.S. 657 (1965). The court in Noerr-Pennington established that no violation of the Sherman Act wil be found where parties have merely attempted to influence the enforcement or passage of laws, even if the result is to injure their competitors. 365 U. S. at 139. 392. Commission counsel contends that the fact that GM and NADA have discussed crash parts distribution for many years (CCPF 464 Initial Deision 281, 282) indicates that settlement proposals to the FTC were not a factor, and therefore Noerr-Pennington is not applicable (CCRB at 117). Obviously GM and its dealers, in the relationship of franchisor and franchisees, have many common interests which require communications concerning a great many matters, not the least of which, no doubt, is crash parts. The fact that they have had discussions on this topic is insufficient evidence of a conspiracy to convince me that a violation of the antitrust laws resulted. GM' submission of proposals to the FTC which called for sales by GM to IBSs, which the dealers oppose very strongly (Finding 300) is further evidence that no conspiracy existed.

393. The importunings of GM and its dealers addressed to the FTC are protected by the Noerr-Pennington doctrine. Mt. Hood 607 (9th Cir. 1977), vacatedStages, Inc. v. Greyhound Corp. 555 F.2d and remanded on other grounds, 437 U.S. 322 (1978). Other cases cited by Commission counsel limiting this doctrine are inapposite. On the basis of the evidence here, GM and its dealers should not be faulted insofar as their actions oriented toward mutual interests are concerned.

Is The Wholesale Compensation Plan GM Uses Illegal? 394. The answer to the question in the caption is "Yes . GM' wholesale compensation plan is unfair. tiThe essence of unfairness in an exclusive arrangement as a marketing tactic is the actual foreclosure of business rivals from consuming markets, thereby compete on even terms. (emphasisdenying them opportunity to added)." Attorney General's Committee Report, supra at 148. The key " inwords are "denying them opportunity to compete on even terms other words, discrimination. The quotation is not on all fours with the instant situation because the business rivals in question are IBSs those of the dealers rather than of GM. Even so, the "denial" suffer stems from GM's wholesale compensation plan and thus violates Section 5 FTCA. (80) 395. Counsel to GM takes the position that the Complaint does not charge that the wholesale compensation plan brings about discrimination. I do not agee. Paragraph 13, in essence, sets forth a charge that IBSs frequently pay more for new service GM crash parts than do GM dealers, and paragraph 16(e) alleges that GM uses disparate" wholesaling incentives in distributing such parts. In addition, the "Notice of Contemplated Relief' on the last page of the Complaint should have alerted counsel to GM to the fact that if the Commission concluded from the record developed that GM is in .. .. ..

Initial Decision 99 F. violation of Section 5, it might order such relief as is supported by the record and is necessary and appropriate, including, but not limited to the "Notice Order 396. Furthermore, there is no requirement that a complaint in an administrative proceeding must enumerate precisely every event to which an administrative law judge may initially or the Commission may thereafter attach significance. The purpose of the administrative complaint is to give the responding party notice of the charges against him. 1 Davis Administrative Law Treatise Sections 8.0 05 and cases cited therein. The complaint is adequate if " the one proceeded against be reasonably apprised of the issues in controversy, and any such notice is adequate in the absence of a showing that a party was misled. Cella v. United States, 208 F. 783 789 (7th Cir. 1953), cert. denied 347 U.s. 1016 (1954); Swift Co. v. United States, 393 F.2d 247, 252 (7th Cir. 1968). 397. As the Commission case unfolds there must be " reasonable opportunity to know the claims of the opposing party and to meet them. Morgan v. United States, 304 U. S. 1, 18 (1938). There is no question here whether GM knew the claims of Commission counselor that GM had ample opportunity to meet them. GM did know and did address them.

398. The plan GM uses for distributing crash parts, regardless of whether a part is eligible for wholesale compensation, results in dealers as a class paying GM less for a crash part than IBSs as a class pay the dealer for the part, even though they may be competitors in repairing crash damaged vehicles. Dealers' final costs for parts vary because: (1) the part may be used in the dealer s body shop rather than resold (Finding 77); (2) a dealer may stock parts and have certain expenses which another dealer does not; or (3) a dealer may not stock and he will have a different set of expenses (Finding 79). No doubt there are other significant factors, but without having detailed information regarding the operations of particular dealers and IBSs it is not possible to determine precisely what the expenses to each in particular competitive situations are, and what the final cost of a particular part is. Such precise data is not needed to support the conclusion that generally IBSs pay more than GM dealers and (81) otherwise are disadvantaged and discriminated against. 399. Dealers buy for 40% off GM's list price (Finding 119). They can get an additional discount depending on whether they order a part on the PAD (Findings 120 and 190(1)) or are entitled to a rebate because they resold the part in a way qualifying for wholesale compensation (Finding 76). They also may have freight prepaid by 464 Initial Decision GM on routine orders (Finding 190(1)(2)(3)) and inevitably enjoy faster delivery service than do IBSs.

400. In contrast, IBSs buy parts at 25% to 40% off GM's list price, averaging about 32% (Finding 124). They cannot get the additional (PAD) discount for the way in which they order or a rebate for a particular type of resale. They often must pay for delivery and there is an inherent delay factor in the time required for delivery (Finding 131).

401. To the extent that a dealer actually performs real wholesaling functions in doing business with IBSs maintaining an inventory, taking orders, ordering from GM, receiving parts and delivering them, extending credit, etc., a particular dealer s final cost for a part may be increased or decreased, but there is very little uniformity amongst dealers in performance of these functions. Some stock, many don t (Finding 102). Some have special order taking and placing facilities, others don t. (Finding 132). Thus, there may be significant variations in the ultimate !lcost" of a part to a particular dealer, but there is no variation in the price he pays GM: 40% off list plus on routine orders the PAD discount, freight prepaid, and the rebate for wholesaling in accordance with the plan (Finding 119, 120 190(1)(2)(3) and 77). The result is that dealers, in general, are favored and IBSs are discriminated against.

402. The plan also discriminates among GM dealers, because those who sell or who may wish to sell a crash part for a make of GM automobile other than the one or more for which they are franchised (Finding 104) are not eligible to claim wholesale compensation. No doubt this prohibition inhibits most dealers from selling crash parts across franchise lines. If a dealer does cross his franchise liners) his cost for the crash part is higher than the cost to any competing franchised dealer. (Finding 77).

403. Discrimination of this sort insofar as the antitrust laws are concerned the Sherman and Clayton Acts, is not clearly ilegal. The prevention of such disfavoring/discriminating, however, is within the spirit of those laws, because it creates an undue impediment to the competition IBSs can offer to GM dealers and to competition among some GM dealers. The wholesale compensation plan makes it impossible for IBSs, as a class, to be on equal footing with GM dealers as (82)a class. The same is true of franchise-crossing dealers and those GM dealers against whom they compete. Such situations may be appraised by the Commission and ultimately by the courts simply as being fair or unfair. FT v. Gratz 253 U.S. 421 427 (1920).

404. The Commission has broad powers to declare trade practices , Initial Deision 99 F.

unfair. That power is particularly well established with regard to trade practices which conflict with the basic policies of the Sherman and Clayton Acts, even though such practices may not actually violate those laws. FT v. Brown Shoe Co. 384 U.S. 316, 321 (1965). 405. In its Statement of Basis and Purpose of Trade Regulation Rule 408 Unfair or Deceptive Advertising and Labeling of Cigarettes In Relation to the Health Hazards of Smoking, the Commission described the factors it would consider in evaluating what constitutes unfair or deceptive practices. These, in essence, were: (1) whether the practice, regardless of legality, offends public policy, is within the penumbra of some common-law, statutory, or other established idea of fairness; (2) whether the practice is immoral unethical, oppressive or unscrupulous; and (3) whether it causes substantial injury to consumers, or competitors or other businessmen. 29 Fed. Reg. 8324, 8355 (1964). The wholesale compensation plan GM uses, among other things, (1) is within the pneumbra of trade regulation and antitrust law to the effect that commercial discrimination is not fair; (2) is unduly oppressive to IBSs; (3) injures consumers by contributing to rises in price; (4) injures IBSs by discriminating against them, and (5) by paying wholesale compensation to nonstocking dealers, who do not perform that warehousing function in distributing crash part, discriminates against stocking dealers. See also F. T. e. v. Sperry Hutchinson Co. 405 U.S. 233 239-245 (1972) and Gratz, supra, 253 U.S. 421, 427. GM's No Public Interest Defense 406. In its Answer to the Complaint, counsel to GM denied that the Commission had reason to believe that use of GM's crash parts distribution system is a violation of the law and denied that these proceedings were in the public interest (Answer, par. I). The Commission has said, however, that such defenses go to the mental processes of the Commissioners and will not be reviewed by the court. Once the Commission has resolved these questions and issued a complaint, the issue to be litigated is not the adequacy of the Commission s pre-complaint information or the dilgence of its study of the material in question but whether the alleged violation has in fact occurred. Exxon Corporation 83 F. C. 1759 at 1760 (1974) (Order denying respondent' s motions for reconsideration of Commission s prior denial of respondents' motion to dismiss complaint). 407. It also should be mentioned that in Herbert R. Gibson, Sr., 90 C. 275 (1977) (Order denying (83Jrespondent's motion to dismiss for lack of public interest), the Commission held that administrative . . . ), 464 Initial Decision law judges lack authority to rule on " questions pertaining to the Commission s exercise of administrative discretion." In that case, the existence of public interest was questioned as part of a motion to dismiss. The Exxon decision, noted above, and a number of other cases were cited.

The Commission s Power To Charge Only One Respondent and To Effect Remedies 408. The fact that the Commission chose to issue a complaint only against GM when all the other vehicle manufacturers doing business in the United States use the same system (Finding 34) is not significant. The Commission has the power to enter an order against one firm that is using an industry-wide ilegal trade practice. F. T. v. Universal Rundle Corp., 387 U. S. 244 (1967); Standard Oil Co. of California v. United States 337 U.S. 293 (1949); Moog Industries Inc. v. F. T.e., 355 U.s. 411 (1958).

409. There have been no instances where a Commission order has been set aside simply because it was directed against a single violator in the face of industry-wide violations. Rabiner Jontow Inc. v. T.C. 386 F.2d 667, 669 (2d Cir. (1967)). However, the Commission s orders must serve a remedial and not a punitive purpose. F.T. e. v. Ruberoid Co. 343 U.S. 470 (1952); Niresk Industries, Inc. v. F. T. e., 278 F.2d 337 (7th Cir. cert. denied 364 U.S. 883 (1960). Further the Commission may not issue oders which would arbitrarily destroy one of many violators in the market. Universalof theRundle, supra 387 U.S. at 251. A "reasonable evaluation" competitive situation must be made to ascertain whether a particular order would be contrary to the purpose of the laws sought to be enforced. 387 U.S. at 251-52.

CONCLUSIONS 410. GM is engaged in commerce and affects commerce as commerce" is defined in the Federal Trade Commission Act. 411. GM has not abused its monopoly in, and monopoly power over, the distribution of new GM crash parts, as defined in the Complaint.

412. GM's refusal to sell crash parts directly to anyone other than GM dealers is lawful.

413. Due to the disfavoring of and discrimination against IBSs and some GM dealers, and in violation of Section 5 of the Federal Trade Commission Act, the wholesale compensation plan GM uses Initial Decision 99 F. unfairly hinders and injures competition in the distribution of new GM crash parts. (84) THE REMEDY 414. The order attached is intended to be remedial and is not contrary to the purpose of the Federal Trade Commission Act. It is not punitive, certainly wil not destroy GM, and I believe it to be just. It is in harmony with both the Federal Trade Commission Act and the antitrust laws.

415. Latitude is necessary in framing orders for ". . . the Commission alone is empowered to develop that enforcement policy best calculated to achieve the ends contemplated by Congress and to allocate its available funds and personnel in such a way as to execute its policy effciently and economically. Moog Industries, supra, 355 V.s. at 413. But the latitude and the broad authority which the Commission has in framing orders must be used with great care to achieve maximally and effectively ends to the interest of the public (15 V. C. 45(b)) which the Congress contemplated when the FTC Act was passed. The circumstances under which the wholesale compensation plan GM uses was implemented (Findings 29-55) and the unforeseen results of its implementation (Finding 417 below) emphasize the need for great care.

416. The evidence establishes that GM's use of the wholesale compensation plan must be stopped because it ilegally discriminates against IBSs and discriminates between stocking and nonstocking GM dealers and may do so with regard to wholesaling GM dealers who cross franchise lines.

417. The plan has not achieved the price parity between GM dealers and IBSs which was the objective when it was made applicable to crash parts in 1968 (Finding 54). Most importantly has increased costs to consumers and has not lowered prices to IBSs (Finding 54). It is expensive to administer, requires auditing, is susceptible to fraud, and does not reward volume buyers for the costs they save GM (Tr. 14002-03; 14013-14) (CCPF 307, 310-14, 352). 418. Although these adverse effects are clear, there is insufficient evidence in this record regarding the provisions best suited for a nondiscriminatory plan for distribution of crash parts. I am not convinced that the relief Commission counsel advocates in the Complaint, (in essence, opening GM's warehouses to everyone except individual owners of vehicles) would be the proper remedy. For example, allowing anyone to buy from GM warehouses at nondiscriminatory prices may cause more problems than it solves. 464 Initial Decision On the other hand it might solve all the problems. Or, the definitions of crash parts and components and applicabilty of the plan may be too limited. It may be unwise to define "auto" and "light trucks" Perhaps crash parts for more vehicles should be included in the plan or the definitions Commission counsel proposes may be too restrictive. (85)There mayor may not be a justification for continuing to categorize crash parts into compensable and noncompensable groupings. No doubt there are other possible problems. 419. The order below, in practical effect, calls for a GM-Commission staff, cooperatively devised plan to bring about compliance with the Federal Trade Commission Act. It wil enable GM, which has the best information about its needs in crash parts distribution, and the Commission and its staff, to see to it that an effective and lawful plan is devised. There was Commission staff involvement in GM' adoption of the wholesale compensation plan it uses now (Findings 29-55). However, the plan in operation has proven to be neither lawful nor to the interest of the public. Commission oversight with the benefit of the record of an adjudicative hearing is the added factor which was not present before and wil lead to the necessary changes.

420. Accordingly, and pursuant to authority contained in Commission Rules 3.42(c) and 3.51(b) the following order is entered. (86) ORDER It is ordered That GM is to submit a detailed report to the Commission within 30 days of the date on which this Order becomes final describing a nondiscriminatory plan which it proposes to use for distributing new GM crash parts.

It is further ordered That within 30 days of the date of approval by the Commission of the new plan GM is to implement its use. It is further ordered That annually within ten days prior to the anniversary of the date this Order becomes final, for a period of five years, GM is to submit in writing to the Commission a report setting forth in detail the manner of GM's compliance with this Order. Opinion 99 F.

It is further ordered That Commission approval of the Plan does not relieve GM of the obligation to comply fully, and in the future and with respet to any changes in practice which GM may from time to time implement in connection with its sales of servce GM crash part, with all of the requirements of Paragaph I of this Order. After the Commission has approved it, GM is not to change any of the terms or conditions of sale set out in the Plan (excluding prices) without first (a) giving 30 days' prior written notice of each such change to the Commission, (b) to all GM customers who purchase new GM crash part, and (c) publishing a description of each such change in (87)Automotive News or a similar publication(s) with wide circulation among independent wholesalers and independent body shops.

OPINION OF THE COMMISSION By BAILEY Commissioner:

This case presents the question of whether General Motors Corporation (GM) has violated Section 5 of the Federal Trade Commission Act, as amended (15 V. C. 45), by its use of a selective distribution system for new crash parts for GM automobiles and light trucks (hereinafter "crash part"). Crash parts, which wil be described in more detail below, are a type of automobile replacement part: true to their name, they are used to restore those part of the car body most commonly damaged in accidents, such as fenders doors, and hoods. GM is the sole source for new crash parts compatible with GM vehicles; it has a long-standing policy of sellng its new crash parts only to its franchised dealers. As a result of this policy, any person who wants a new GM crash part for installation resale purposes must get it from a GM franchised dealer. Particularly disadvantaged by GM' s policy are the independent body shops (IBS), specialists in automobile body work, whose business depends upon a supply of crash part. IBS are supplied new GM crash part by GM-franchised dealers, who are also the IBS' competitors for collision repair business. Despite GM' s effort to encourage its dealers to resell crash parts to the IBS at cost, the record clearly shows that IBS as a class pay more for crash parts than do their rival dealer-installers. (2) GM' s selective distribution System for crash part also disadvantages another class of businesses, but in this case the disadvantage is purely theoretical. The businesses in question are the independent 464 Opinion wholesale parts distributors (IWs) who currently sell automobile replacement parts which are available from more than one manufac turer. Examples of these multisource parts are spark plugs, shocks and glass. There is some evidence in the record that IWs would be interested in wholesaling GM crash part, thereby introducing competition to a business which is presently the sole domain of GM. This appeal reaches us with a lengthy history: three investigations from the mid-1960' s until issuance of the Complaint on March 22 1976, two years of discovery, and 82 days of hearings which produced a bulky record in no little state of disarray. ' In any proceeding as long and involved as this one, it is possible to pick away at the details and find numerous flaws and inconsistencies. Both respondent and complaint counsel have been quick to explore everyone. But in appellate review our function is broader. We look to the entire record.

The primary diffculty with this case is not in perceiving that GM' selective distribution system disadvantages IBS in the business of installng crash parts or disadvantages IWs in the business of wholesaling crash part. The problem is, rather, whether these disadvantages translate into the sort of injury to competition which is cognizable under the antitrust laws. The diffculty in arriving at a determination on that question can be seen in the fact that complaint counsel argued at trial and on appeal that the system amounts both to a vertical and a horizontal boycott, several types of attempted monopoly, and an abuse of monopoly power. The Administrative Law Judge (ALJ) rejected all of complaint counsel' s arguments as to the ilegality (3Jof the system as a whole. (IDF 380, 411 412)' However, he did sua sponte find that one component of the , The rerd contaIU approximawJy 8 40 exhibits a.nd the testimony of 84 witnel, running to appro1timatey 1628 paes of tre.ript. It doo not contain any BOrt of document list, much les any table referencing tetimoflY authenticating, identifying or explaining II given document. Moreover, many documents were reived piecmeal-part fl respondent s cJthibit, part lI complaint counsel's exhibit, sometimes part Il theALJ' s exhibit-without any clue Il to where the remaining portions of an exhibit could be found. 'I disarray Cr'te no little problem in reviewi this ca. . The following abbrevitiolU wil be ua in this opinion: IDF - Initial Deision Finwng Number 'r. - Transcript Page Number, preed by witness' name AJX - Adminisrative Law JUdge B EJIhibit Number Complaint Counsel' Exhibit Number CAd - Complaint Counsel'B AdmisionB CPF - Complaint Counsel'B Propo Finding CRF - Complaint CoulUl'B Reply Finding - Complaint CoulUl'B Pool Tral Brief CRB - Compla.int CoulUl'a Poot Tr Reply Brief CAB - Compla.int CoulUl'a Appeal Brief CAAB Complaint CoulUl' B Appellate AnBwerig Brief CARR - Complaint CoulUl'B Appellate Reply Brief CASB - Complaint Count!l'B Appellate Supplemental Brief on theReuben /JnnelkydeciBion (Continue Opinion 99 F.

crash part distribution system, known as the Wholesale Compensation Plan (WCP), was unfair and a violation of Section 5. (IDF 413) (4) We affrm the ALJ's reasoning and conclusions oflaw that GM has neither attempted to monopolize distribution of crash parts, nor engaged in vertical or horizontal boycotts effecting limitations on crash parts distribution. Like the ALJ, we conclude that GM has not abused its monopoly power over crash parts distribution, but we reach our holding by a different and more complex process. The ALJ interpreted the precedents as giving every supplier, even one having monopoly power in the relevant market, an unassailable right to refuse to deal with willng customers. Since he found that GM has no predatory or monopolistic purpose in its choice of distribution system, his analysis began and ended at this point. Our approach is to determine if this case presents one of those rare situations where a supplier with monopoly power over an essential product has a duty to deal with all customers. This requires a rule of reason analysis which weighs the injury to competition against the supplier business justification for its actions. In making such an analysis our first step is to determine the extent of harm caused by the refusal to deal. Only substantial injury to competition compels us to look further and assess the business justification for the refusal to deal. The economic self-interest of the monopolist is always an important consideration, but becomes particularly compelling when the harm shown to competition is not clearly and directly traceable to the refusal to deal, or can just barely be characterized as substantial. In this case, although we find injury to independent bodyshops caused by the effects of General Motors' selective distribution system for - Respondent's Exhibit Number RAd - Respondent's AdmiBiona RPF - Respondent's Propo Finding HRI" - Respondent's Reply Finding - Respondent's POtt Trlll Brief RRB - Repondnet's POtt Tral Reply Brief - Respondent's Appeal Brief RAB - Respondent's Appellate Answering Brief RARD - Respondent's Appellate Reply Brief RASB - Respondent's Appellate Supplemental Brief on the Reuben Dmnelky decision INX - Intervenor NADA's Exhibit Number INPF - Intervenor NADA's Propo Finding INRI - Intervenor NADA's Reply Finding INB - Intervenor NADA's Post Tral Brief INRB - Intervenor NADA's Poot Tral Reply Brief lab - Intervenor AS' s Poot Tris.! Brief IAN - Intervnor AS's Post Tral Reply Brief - Intervenor AS' s Appeal Brief IASB - Intervenor AS' s Appellate Supplemental Brief on the Reuben Dunnelley decision ;...

GENERAL MOTORS CORP.

464 Opinion crash parts, we cannot say that that system causes any enduring weakness to the IBS as a class of competitors or that GM' s choice of distribution system is arbitrary or without substantial business justification. Finally, we conclude that the WCP merely reflects and does not add in any way to the inequities which the selective distribution system imposes on the IBS and IWs. We reserve the ALJ in his finding that the WCP, as distinguished from the selective distribution system, injures competition in the distribution of new GM crash parts. Accordingly, the case is dismissed. We adopt such of the ALJ's findings and conclusions as are consistent with the findings and conclusions set forth in this opinion. (5) I. BACKGROUND A. General Motors Corporation Respondent General Motors Corporation needs little introduction. Since its organization in the early years of this century, it has grown to be the largest manufacturer of automobiles and light trucks in the United States. (IDF 56, 65) In addition, GM manufacturers a variety of products within the automotive and transportation fields; but the primary interest of both GM and its franchised dealers is in selling cars and trucks. (IDF 57, 58) In 1976, GM-manufactured automobiles accounted for approximately 45.5% of total U.S. automobile registrations and GM trucks for 42% of total truck registrations. (IDF 62) GM manufactures and sells in the U.S. five "lines" or principal makes of automobiles: Chevrolet, Pontiac, Oldsmobile, Buick, and Cadillac. (IDF 57) There are 12 different body sizes available for GMmanufactured automobiles as a whole, and within each line a multiplicity of body models, which often change from year to year. (IDF 62) This stylistic variety significantly influences both the demand for new crash parts and the structure of the distribution system which answers that demand.

Crash Parts Crash parts are a type of automobile replacement part. They constitute the cosmetic, visible, outer parts of a car which give it its distinctive style, and are expected to last the lifetime of the car under normal conditions. They are generally replaced only as a result of collsion damage. They are therefore distinguishable from electrical or mechanical parts which are installed in the internal functional system of the car and, as a result of either wear or failure . , .

558 FEDERAL TRAE COMMISSION DECISIONS Opinion 99 F.

are replaced on a maintenance basis. (IDF 145) The Complaint lists as crash part:

anyone . or all of the following products: fenders, griles, bumpers, hoo; deck lids, doors, quartr panels. rear end pa.nels, rocker . panelsi lamp asemblies, wheel opeIiingpanels, fenders and rear end caps, tal gates, radiator support and shrouds, and mouldings; including inner and outer panels and all components of these products as well as all part necessry to attach the aforesad to the boies of automobiles and light trucks. (Complaint, Par. l(d)) (6) Crash part are non-standard part tailored to fit specific cars, by model and year. Thus a crash part of a Ford car wil not fit a GM car; nor are crash part generally interchangeable among GM automotive lines; nor among different models within a line; nor . even, at times, between different years of the same model. (IDF 71, 144) Small wonder, then, that the GM Part Division carries about 32 000 crash part numbers. (IDF 167) Crash part as a class are bulky and require considerably more space for storage than mechanical or electric replacement part. Crash parts also require especially careful treatment during distribution as they are easily marred and hard to handle. (IDF 74) All GM crash part are produced, either by GM or by independent manufacturers to whom GM has subcontracted the work, from tooling owned by GM. (IDF 68) Crash part sales, although only a small portion of GM's total revenue, are significant: GM computed the gross dollar value of domestic shipments in 1975 as approximately $314 milion (CX 7407A) and complaint counsel's undisputed approximately $548.8calculations for 1977 produced a figure of milion. (CPF 15, 46, 306) Although GM has made no effort to inhibit others from entering the crash part manufacturing business, none have done so. (IDF 151) This requires some explanation. The total demand for crash parts is high. (IDF 152) Demand for individual crash part is fairly inelastic, being stimulated by the vagaries of collisions. (Benston Tr. 16070; Murphy Tr. 10286) The demand for each individual part is generally quite modest. This is due to the low probabilty that a vehicle wil require replacement of a particular crash part during its lifetime. (IDF 152, 153) GM usually start production of a half year supply of crash part for automobiles before introducing the new model, and keeps suffcient inventories of crash part for each model for 7 to 12 years. (IDF 80, 84) It is not economically feasible for GM to )foduce tota anticipated crash part needs in one production run ouch less store such a mass of part. (IDF 154) Thus, the bulk of eplacement crash parts are built over the lifetime of the vehicle as 'warranted by inventory needs and warehouse space. This process GENERAL MOTORS CORP. 559 464 Opiniori requires that dies used in manufacture of new cars be retrieved steam cleaned, reconditioned-in some cases partially rebuilt-and then insertd into presses to run the required supply of crash part. This is an expensive process because the runs are short-term and the technology is labor intensive. (IDF 156) (7) Therefore, expensive as it is for GM, the process would be prohibitively expensive for anyone else. Tooling to manufacture GM crash part can run to tens of milions of dollars. (IDF 149) However the same dies which are required for service part are used in the production of original equipment, enabling GM to spread the costs associated with crash part manufacture. (By GM's estimate, crash parts manufacturing costs amount to less than 15% of total tooling cost). (ALJX9T) It is not diffcult .to understand why other manufacturers have shown no inclination to incur such large tooling costs simply to produce crash part. In this market, GM's ownership of the dies used in the manufacture of new cars gives it a wholly natural and decisive competitive advantage over potential manufacturers of crash parts. (IDF 148) In sum, considerations of.. scale economies and demand, . rather than monopolizing conduct, explain the lack of competition in crash part manufacturing.

C. General Motors Part Distribution Systems General Motors has two systems for distributing the automotive part which it manufactures. (IDF 157) About 60 000 "maintenance typ" part, which can be used on both GM and non-GM cars, are sold to the independent aftermarket by GM's AC-Delco division. (IDF 158, 183-185) About 300 00 sheet metal and engine parts, which are applicable solely to GM cars, are sold exclusively to GM-franchised dealers by the General Motors Part Division (GMPD). (IDF 158, 167) Crash part, with only a very few exceptions, are not AC Delco part and are distributed solely through GMPD. (IDF 158, 165, 186) To evaluate GM's business rationale for its choice of crash parts distribution system, the following points must be made. Of the total number of part distributed by GMPD only 32 000 (approximately 10%) are crash part. (IDF 167). GMPD makes no distinction between crash part and all the other sole source part which it distributes. (Id. Moreover, all U.S. auto manufacturers and all foreign automobile companies that have been sellng cars in the U. since the early 1960's distribute their crash part in the same way that GM distributes its crash parts. (lDF 94) GMPD seems to specialize in the distribution of low demand, slow moving part. 75% of its total inventory fits this description (IDF 175) as do from 65% to , .

560 FEDERAL TRAE COMMISSION DECISIONS Opinion 99 F.

77% of all crash part, depending upon which GM study one reads. (8J(Compare RX 20 with CX 7006D) As we shall see many GM franchised dealers wholesale crash part, but they tend to inventory only the fastest-moving part applicable to their franchise line. Accordingly, most IBS COIn plaints concern the difficulty of getting the low-demand, slow-moving parts, which must be specially ordered by the GM dealer from GMPD inventory and routed through the GM dealer for delivery to the IBS. (IDF 177. 178) D. Crash Part Installers: GM Franchised Dealers and the Independent Body Shops The complexities of actual and potential systems for wholesaling crash parts are more easily understood after a brief review of the retail end of the distribution chain. Almost all body repair work on GM cars and trucks is done by GM dealers or the independent body shops. (lDF 100) In 1976 there were approximately 12 000 car and truck dealerships franchised by GM in the United States. (IDF 2) The primary business of these dealerships, like that of GM, is sellng cars and trucks. (lDF 58) Each dealership has ready access to the crash part applicable to its line, and at least 80% find it expedient to install crash parts as part of the ongoing service they offer on the cars they sell. (IDF 59, 69, 100; Bentson Tr. 15770) They do so for two reasons. First, body work on cars, including crash part installation can be a profiable sideline. (Perkins Tr. 9916) Second, availability of parts and service, in an incondite way, affects a customer s decision to buy a new car. While the record contains no evidence that prospective new car purchasers specifically ask about crash part availabilty, much less cost ' it does support the general proposition that dealers feel that the reputation oftheir body shops helps make a sale. (Perkins Tr. 9914; Bentson Tr. 15751- , 15793) GM has apparently also (9Jconcluded that parts availability can affect new car sales, but that the precise correlat.ion is peculiar to each dealership. The standard GM franchise agreement requires the dealer to carry in stock parts and accessories Hadequate to meet customer demands," but specifies neither dollar amounts nor number of items to be inventoried. (lDF 95) In 1972 there were over 32 000 IBS in the United States. (RX 38 39) IBS are generally smaller operations than GM dealer body shops often being one or two man operations. (IDF 264) While GM dealers tend to specialize in repairing the models they sell, IBS do body work on all types of vehicles. (IDF 100) The record does not show what . Moat car buyers count upon insurance to CQver damage expenBeIl, and 60 devote little thought to future repair COW at timeofpurchss. (IDF 60) GENERAL MOTORS CORP. 561 464 Opinion portion of IBS revenues are provided by sales and installation of GM IBS purchase their crash parts; our rough estimate is around 20%.4 new GM crash part from GM dealers, generally for more than these dealers pay GM for the identical parts. It was IBS complaints about the wholesale cost discrepancy on crash parts which launched this case. (IDF 29) Both the IBS and GM dealers were represented in this proceeding. The Automotive Service Counsel, which had over 2 000 IBS members in 1978, fied briefs. (IDF 15) The National Automobile Dealers Association (NADA) to which 70% of all GM dealers belong, participated in the trial and fied briefs. (IDF 14) GM Franchised Dealers as Crash Parts Wholesalers; the Wholesale Compensation Plan GM dealers, as far as the record shows, have always been free to sell the crash part which they purchase from GM. GM has not and does not attempt to control its dealers' crash part sales with respect to price, customer or territory. (IDF 198) GM does suggest wholesale prices for crash parts. (IDF 197) GM has intruded upon its dealers' wholesaling freedom to the extent of providing incentives for the sale of crash parts to IBS. Under the Wholesale Compensation Plan (WCP) adopted in 1968 GM rebates voluntarily to participating dealers (10Ja percentage of the cost of crash parts sold to IBS.' The purpose of this plan which was worked out after discussions with the Commission staff and adopted also by Ford and Chrysler (ALJX 17-5) was to overcome the crash parts wholesale price disadvantage faced by the IBS. Under the plan, dealers are supposed to wholesale crash parts at cost making their profits wholly from the rebate, and thus putting the IBS at parity with GM dealers on crash part cost. (IDF 34, 76) Unfortunately, many GM dealers eschew such altruism, seeing in the plan an opportunity for "double dipping taking the WCP rebate from GM while sellng crash parts at a markup to the IBS. The extent and effect of the wholesale cost variance between dealers and IBS wil be explored in detail later in this opinion. For now it is important to note that GM cannot force its dealers to wholesale crash parts at cost, or any other price, without laying itself open to a charge of resale price maintenance. Therefore, the dealers exploitation of the WCP is of the same nature as their basic freedom to dispose of goods which they own in any matter they see fit. It does .' in(ro note 74.

. The tehnics.! in and outs of the WCP and dealer orderig method for cr.lh part Ille excengly complex. They will be highlighte WI nec in th opinion, but the reader is reference to IDFs 75-79, 82, 104-112, 100 196 for II complete desription of the plan l! operation. 562 FEDERAL TRADE COMMSSION DECISIONS Opmion 99 F.

not interfere with a dealer s freedom to decide whether .or not to wholesale. (IDF 103) There is a conflct in the record ""to the number of GM dealers who engage in "meaningful" wholesaling," but no indication that the WCP has changed this percentage in any way, or altered the geographic dispersion of crash part throughout the United States. In short, the WCP did not introduce into GM' crash part distribution system any fundamental new disadvantages to the IBS.

II. LEGAL ANALYSIS Conspiracy Theories Complaint counsel argue that GM's exclusion ofthe IBS and IWS from its crash parts distribution system is the result of a conspiracy betweenGM and its dealers. That conspiracy, Janus-like, is alleged to be either horizontal, with GM a willng co-onspirator, or vertical with GM a coerced partner. . Thus, before considering whether any ageement exists we must determine whether GM and the GMdealers share the same niche in the crash part distribution chain. (11 J L Horizontal Theories (a) Presence through the Motors Holding Division The horizontal conspiracy theory is easily disposed of. It is uncontested that currently' GM's only formal presence at the automobile and light truck dealership level (and consequently at the crash part installation and retailng level) is through its Motors Holding Division (MHD). (IDF 87) Since 1929 the MHD has administered a Dealer Investment Plan under which it provides interim capital financing for a small number of new dealerships. Under this plan the prospective dealer is required to make an initial investment of at least 25% of the total required capital by purchase of non-voting stock in a dealership company. MHD provides the balance by purchase of voting stock and issuance of long-term notes. The dealer is paid a salary and shares equally in the dealership s profits on a per-share basis. The dealer conducts the day-tQ-ay operations of the dealership. (IDF 91) While . &e infro note 91.

, In 1974 GM owned and opera.te a very smal Dumber (23) of automobile dealership!. (Md. 763) Thee were pha out by 1976. (!F 87) Hiatoricaly, of COllre, none of the Big Th auto manufacturers ha ever BerioUfly attempted to util the agency form of diribution aB a maketin device. Se gerurolly &:hntt, Antitrut wid Distribution Problema in Tiht OlipolieS; The Automobile Indusry, 24 Chain L-J. 849 871 (1973); E. Cray. Chl"lde ColoUB: General Mota\" and Its Ties 30 (198). );

u.ll'1J:.rftL IV.UIU.! \Au,r.

464 Opinion MHD owns all the voting stock in the dealerships which it finances there is no evidence in the record that it uses its voting power to become actively involved in the daily management and direction these corporations. Cf Chisholm Bros. Farm Equip. Co. v. International Harvestor Co. 498 F.2d 1137 , 1142-1143 (9th Cir. 1974); cert. denied 419 U.S. 1023 (1974); United States v. Sealy, Inc. 388 U. 350, 353 (1967). Thus, as regards any given MHD dealership, GM functions as a financier rather than an operator. (12) Moreover, under the Dealer Investment Plan the dealer has the right and is expected to increase his capital stock interest by purchasing the MHD-held stock. (IDF 91; CX 7029G-E) Between 1970 and 1977 the average length of MHD financing for a dealership was 5 to 6 years. (CX 34) Consequently, the MHD interest in any dealership is temporary.

Nor does the total number of dealerships operating under the Dealer Investment Plan change the nature or degree of GM' involvement with crash parts installation and retailng. There is not the slightest suggestion in the record that GM's crash parts policies were shaped by a desire to benefit the MHD dealerships, and it would be surprising if there were. Between 1929 and 1964 there were only 1 139 MHD dealerships. (IDF 91) The record shows that the MHD dealerships never accounted for more than 3% of the total GM dealerships between 1964 and 1977 ' and that their number has declined since 1970.' There is also no evidence that the MHD dealerships conduct an abnormally large portion of the overall dealer bodyshop business. To the contrary: in 1977, MHD dealerships accounted for an estimated 4.3% of total GM dealer wholesale parts sales and an estimated 5.3% total GM dealer body shop sales. (RX 34; CX 38, 40) In determining whether a business relationship is horizontal or vertical we must useek the central substance of the situation, not its periphery. United States v. Sealy, Inc. 388 U.S. 350, 353 (1967). In central substance, GM's relation to its dealers is that of franchisor and supplier; it is simply not suffciently integrated forward to be clasified as their competitor. " (13) . The ratio ofMHD dealerships to total dealerships W8., for 196, 30:13 395 (2.3%); for 1974, 379:11 894 (3%), for 1977, 34:11 66 (3%). (R 34; ex 38, 40, 2079C) MHD Dealer Equity Rerds and Financial Statements show 449 MHD dealemhips in 1970, 379 in 1974, and 34 in 1977. (R 34) .0 On an apparently similar fa.actual pattem, Ford's tempont.ry ownership interest in an unspeified number ofits franchis dealerships for financing purp w8. held not suffcient to permit a finding that Ford had any sigificat share of the dealer-Jevel market. FLM Collision Part 1m: II.Ford Molar Co. 40 F. Supp. 224, 24 (S. rev d inpart, afrd inpart 543 F.2d 1019, 1030 (2d Cir., 1976), cert. cUnie, 429 U.S. 1097 (1977). Opinion 99 F.

(b) "Symbiotic" relationship Complaint counsel also argue, albeit weakly, that GM has a meaningful presence at the wholesaling and installation level due to a "symbiotic relationship" with its dealers. This relationship amounts to nothing more than the general interest every supplier of consumer goods has in seeing that its products are ultimately sold to the public. To equate this community of interest to functional horizontality does total violence to the normally understood meanings of horizontal and vertical in the distributional context. In conclusion, the AU correctly found that the essentially vertical relationship between GM and its dealers precludes a finding of horizontal conspiracy in this case. (IDF 373-375) Vertical Theory The scenario becomes much more complex under this approach. The issue is not the nature of the relationship between GM and its dealers-it is clearly vertical-but what in fact was their course of boycott at law.conduct and whether it amounts to a vertical Specifically, complaint counsel urge that GM seriously considered two different ways of opening its crash parts distribution system, but that each time dealer pressure, largely through the offices of their national trade association NADA forced GM to change its plans and keep the system closed to all but franchised dealers. Proof of these allegations is extremely delicate, since the denouement involves no clear-cut action by GM, but rather a continuation of its previous course of conduct as evidenced by what it did not offer the FTC in the shifting terms of (14Jthree GM settlement proposals during the ongoing staff investigation which culminated in the issuance of this complaint. Also relevant are the terms of settlement urged by NADA as an interested third party, contemporaneously with GM's second settlement proposal. (CS 7327; AUX 2) The actions of GM and its dealers in the years preceding the 1975- 1976 negotiations with the Commission staff are our only source from which to infer whether GM's settlement proposals were actually dictated by its dealers. Accordingly, that course of conduct I L As note previously the National Automobile Dealers ABiation (NADA) represnta about 70% of all GM dealer" and il an intervenor on behalf of GM in thi ca. Its organiztional structure and that of GM's dealer adviry boies are describe at IDFII 289-291. Throughout thi seion of the opinion "NADA" and "dealers" are us interchangeably.

,. A word of explanation about the multiplicityof citations to a mere throo settlement offers. Unfortunately, the jigw nature of thi record reches its apotheoia in the piecmeal admiBion of thes propo, The First Propol (July 11, 1975) is found acsttered throughout ex 7010, 7012 735; ALX 11, 13C-D, 15. The Second Propo (February 5, 1976) iB found at ex 7012, 735; ALX 13B, 13B, 15. The AL rejec direct proof that GM made a third formal settlement propo. However, the record doe contain indirect proof of this offer, by meall of teimony of both GM representatives and the Director of the Bureau of Competition before tht! Consumer Subcmmitte of the Senate Committe on Commerce, on Man:h 12, 1976. (R 28; ALX 14, 17) , U.D!....nru. U"".L"".L"" ....n Opinion 99 F. T. is chronicled at some length here, as a prologue to application of the relevant legal principles. For clarity s sake, even though the cast of characters and some events are the same, we give two histories: first the line of events relating to a proposal to open GM's warehouses to IBS, which was offered in GM's first settlement proposal, withdrawn in its second, and indirectly reoffered in its third; second, the line of events relating to the mpre elusive "miniwarehouse plan" which was not offered to the FTC in any of the settlement proposals. (a) Unsuccessful Settlement Proposals The critical time period of this history is July 1975 - March 1976. It was during this time that the dealers became seriously concerned that GM might be planning to stop dealing crash parts exclusively to them and start sellng directly to (15)their IBS competitors." The dealers' fears were not unfounded. Almost since its inception in 1968 the Wholesale Compensation Plan was an obvious failure at creating crash parts price parity between dealers and IBS. (IDF 44) GM felt (rightly as it turned out) that unless it could remedy the situation in some other way the FTC was likely to challenge the distribution system as an unfair act under Section 5. (IDF 45) Accordingly, on July 11 , 1975, GM submitted a three-point settlement proposal to the FTC (hereinafter the "First Proposal"). (CX 7010; ALJX 11) It offered (1) To publish the suggested general trade (i. wholesale) price on crash parts as defined by the FTC. ,.

(2) To include subcompacts and trucks in the Wholesale Compensation Plan (WCP), provided the FTC could either persuade or compel GM's competitors, including foreign competition, to do likewise.

(3) To sell to IBS and GM warehouses at dealer price. NADA was given no advance warning of this proposal. (CX 7321; McCarthy Tr. 3560) Dealer reaction was instant and, as GM' counsel has stipulated vehemently" in opposition to Point 3, which came to be known as the "open warehouse" issue. (Tr. 3376) NADA " Even hefon! 1975 dealer' wen! oppo the poibility of change6 in the crah part distribution system but any pre1975 meeting sern consist of general exchllgcs of view-nothi speific enough mark the benning ofa corcive NADA compai or a GM-NADA conspiracy. (McCarthy Tr. 3518; PohllkB Tr. 4690, 4789). Sigifcatly, in thea pre-1975 meeting GM expre sympathy but would not commit the dealer"' position, c:efully keeping open ita optionsfor change. "I know that on more than one ocion Mack Worden lGM's Vice President fof Marketing) would MY; This ia how we fee! about it now, but, of cours, we can chane our minds wmorrw. I remember that." (PohankB Tr. 469B) 14 Th was wallow lB to tell ua GM dealer was cheatin on the WCP by marking up a part he was BUppo to be sellng at cot. (ALX 2C) 464 Opinion organized to fight the proposal. (McCarthy Tr. 3564) During the next four months NADA continually urged (16)its views to the public to GM " and to its members." The members responded by sending a considerable number of letters. . . to General Motors saying that they thought these actions served neither the dealers nor the consumers. "" (Pohanka Tr. 4721) Yet NADA had no confidence that all this activity was paying off. GM appears to have expressed sympathy for the dealers' position (CX 7305), but NADA was unsuccessful in getting a commitment to withdraw (McCarthy Tr. 3518, 3565-6) or even compromise (CX 7305) on the controversial proposal. Indeed, by early December 1975 the President of NADA's GM Line Group, Walter Stilman, believed that GM's position had actually hardened: They have already indicate a willingnes to open the warehouses to independents at a Price as state in their Propoal of July 1975. Thus it is not a question of wil they open them, it is purely a question of at what Price, and it is for sure if the Dealers are to be the Sacrificial Lab they wil not hesi"'te long. (CX 7303C) (17) NADA Executive Vice President McCarthy feared matters might get even worse, that GM might accept an FTC counterproposal to open its warehouses to IWs as well as IBS. (McCarthy Tr. 357"'75). Accordingly, NADA startd working on a new tactic: rather than rely on letters and phone calls to prod GM into recognizing the dealers' plight, it decided to represent dealers itself before the FTC by means of a "4 Point Program" to be considered as an alternative to GM's stil outstanding First Proposal. (McCarthy Tr. 357"'74) (CX 7303A) The program was essentially worked out by December 17, 1975 (McCarthy Tr. 3526) when GM and NADA officials had the last meeting which is documented in this record. As wil be seen, NADA's 4 Point Program submitted to the FTC on February 5, 1976, was IS ex 7301 (July 25, 1975 NADA Pr Releas); ex 736 (11120/75 NADA letter to varous Congmen and Senators).

We were very dist when General Mo\.ra made the offer.. And every opportunity I had to tell General Mowm about that, I told them:' (Pohaa Tr. 4718) I know all the (NADA) councils were fIghtin GM on it, or doing everyhi they could in remmending to them that they back off ofit " (Vernon Tr. 331) ex 7319 (undate Mailgnun, ae 'f. 33, 3373-74); CX 7303A; McCrthy Tr. 35, 35;CX 730; Stilman Tr. 8177-78.

17 ex 7314 (Augt8 1975 circulation to NADA members). .. ex 7310 may be an exaple of the type of letter that W81 wrtten, although it Wil Bent to the FT, not GM. The rerd conl.in DO examples of dealer letters to GM nor any better eatimate of their number tha Pohanka' tetimonycapa. cite above. Thus it il impoihle to evaluate the corcive tone or power of thi pha of NADA' I. The rerd reflects that 108 of their entire wholeain hUBme6 to IWB or the itUurance complliea W81 a far grter threat to dealers tha the apere of the IB achievi price parity in crash pam. There ia even some evidence that NADA felt it could live with openi warehous to IB (81 oppo to IWa) if that were the outer bound of expading GM' a dire distribution ayatem. (CX 730E, I;McChy Tr. 3529) GENERAL MOTORS CORP. 567 Opinion 99 F.

essl:ntially the saml: as GM's second proposed settlement offer (hereinafter "Second Propoal"), also submitted on February 5, 1976. Both eliminate the third point of GM's First Proposal: sales to IBS from GM's warehouses. Therefore, if any ageement took place between GM and NADA on the IBS question, it happened at the December 17, 1975, meeting.

NADA attendees who testified were Frank McCarthy, Executive Vice President and Kevi Tighe, Legislative Counsel, who, however attended the meeting only briefly and sporadically. (Tighe Tr. 9505) Both also made notes of the meeting, which are in evidence asCX 7316 (McCarthy) and CX7324 (Tighe). Also present were, for NADA: Jack Pohanka, President-Elect; Walter Stilman, Chairman of the Industrial Relations Committe's (IRC) GM Line Group; Paul Herzog, Director of Research and Dealership Operations; and Jay Ferrand, Asistat Director of Research; for GM, Michael Meehan Executive in Charge of Servce Part Operation, and Jim Melican attorney. (McCarthy Tr. 3519-20) The purpose of the meeting was to discuss the latest developments in the FTC investigation. (Tighe Tr. 9506) McCarthy, as might be expected, testified that while NADA representatives did tell GM what "we thought would be the proper approach to solve the crash par program (i. the Four Points) they did not ask for GM's opinion of or concurrence in their proposed program. (McCarthy Tr. 3525-27 , 3570-71) (See generally Pohanka Tr. 47 9) (18) Against these statements must be set one sentence from Tighe notes: "What can GM and dealers do together to keep independent distributors out of crash part area????" (CX 7324B) Tighe s explanation of the sentence, constantly repeated in his testimony without change, is that it was a paraphrase of what Melican reported as being the FT' s concern at that time. In other words, Tighe thought he was told that the FT was thinking of isuing a complaint which included a conspiracy count.

(A)corclng to Mr. Melica, as far. as crystal ball gazing, if you will, FT would probably come forth with some form of a complaint which would involve the issue of t its dealers, doing something in the area of conspiring to keep independents out. Tht is why the asterik is at the top20 and it reflects Fr has never answered, and then it picks up later on with Melican s, as I state, analysis and his opinion of what would. perhaps be forthcoming at a later date from the FT. (Tghe Tr. 9518; see also Tr. 95010, 9520-21) McCarthy stated flatly that this portion of Tighe s notes is inaccurate. (T. 3522) U Th is 8. reference to the position of the key sentence on the document, from which both parties attempted to iner 8pew me.

568 federa TRADE COMMISSION DECISIONS 464 Opinion Despite Tighe s uncooperative conduct as a witness we are inclined to accept his explanation for the following reasons. The sentence position.does appear in the context of a summary of the FTC's Tighe s area of expertise was "work on Capitol Hil"; he was unfamilar with the FTC crash part matter which was handled by outside counsel. (Tighe Tr. 9507) That, plus his sporadic attendance of the meeting, could explain his misunderstanding Melican. The record is silent with regard to any actions by either NADA or GM after December 17, 1975, until February 5 1976, when each sent a separate settlement proposal to the FTC." The proposals were later de!Iribed by NADA official Cecil Vernon as "essentially the same."(CX 7353B; Vernon Tr. 3389-90) Both advocated improving crash part distribution by increasing the dealers' opportunities under the WCP: (19) NADA Proposal GM Proposal Increase WCP to 30 % GM states that it has (CX 7327E)" already raised WCP to 30% in October, 1975.

(ALJX 13G) Publish wholesale GM wil publish wholesale prices. of crash parts prices of crash parts (CX 7327F)" (ALJX 13A-B)" Require GM to offer WCP GM wil make WC? available on all models, including on subcompacts and lightduty subcompacts and compacts trucks on a trial (CX 7327G: ALJX 2D)" basis. with the option of withdrawing at the end of six months if principal competitors, including the foreign competition, fail to implement a similar plan. (ALJX 131)24 0' The reord doe not reea the rcalU behind GM's timing of the Send Propor. It doe show tht NADA p!rrived grwing rr preU\" on GM in the new year. and 11 made haate me ita propo before it was to Jute. (McChy 'f. 3574-75; ex 7324B) Th doe not, of cours. explain the fact of contemporaneous nJing. .. NADA reliz that GM iu alrey taken thi step (ALX 2C) but apparently felt that it needed to be ratined by the Ff. The incre to 30% W81 suppo to tae away the preaure overchare to IB. due to inadequate wholese compenstioIJ. (AL 2C) .. This provision aI appered in GM'B Firnt Prpoi. .. Th is a slight variation on the send point in GM's Firat Propor. Opinion 99 F.

NADA Proposal GM Proposal Require GM to offer the GM will pay wholesale WCP across division compensation across division lines. (CX 7327F)25 lines, for sales to IBS" (ALJX 13G) (20) It is obvious that the key differences between GM's First Proposal and these two proposals are the absence of the provision on direct sales by GM to IBS and the appearance of the provision on extending WCP to cross division lines sales. Both February proposals make it clear that broadened WCP is an alternative which negates the need for direct sales to IBS.

(fin lieu of opening up the 27 General Motors Part Division s field warehouses for direct sales to the independent auto body repair shops, General Motors would be willng to agee to broaden the Wholesaling Compensation Plan.. . (ALJX 13G) (emphasis in original) (Olpening the manufacturers' warehouses. . . is also a prime example of governmental overkil, Any dealer overpricing can be simply and effectively by-passed by adoption of points 1 and 2 of NADA's proposal (ALJX 2D) The convergence of timing and content in these two February proposals is certainly remarkable, but before considering whether it shows a conspiracy, we must consider GM's action just a month later. On March 1, 8, and 12, 1976, the Consumer Subcommittee of the Senate Commerce Committee held hearings on the cost of automobile crash parts." In his testimony on March 1, 1976, the Director of the Bureau of Competition described the First and Second Proposals suggesting that the change on the sales to IBS provision was the result of successful dealer pressure on GM. (ALJX 17, Johnson p. 8) In response, by letter of March 5, 1976, to the Director of the Bureau of Competition (CX 7013 rejected" ) and by the March 12, 1976 testimony of (21JMichael Meehan, General Motors added a fourth element to the Second Proposal:

Last week, we added a fourth element to our proposal .. The WCP did not and atil doe not offer rebate on rcaes of crash pam uo on a brand of automobile which the dealer is not franchi to sell. Thus, under thi propo a Pontiac dealer, who previously received wholcae compenation solely for rele of Pontiac crash pam, would be eligble for wholeHe compenation On the sale of Chevrolet, Oldsmobile, Buick and Cadilac crash pam WI well. I. The complete tnmacript of theheari is entered in the rerd aB ALJ 17; portions of the testimony from the hearings are alavaiJable as RX 28, ALX 14. IT ex 7013 is the letter of March 5, 1976, from GM'a General Counsel to Owen Johnn, Director of the Bureau of Competition. Complaint Counsel sought to introduce it, pot trial, in reponae to respondent'a propo replyfindings. By order of September 6, 1979, the AL rejected CX 7013 aB being "unnec'' to his decision. We evaluation of GM's continuing interaction with NADA.dis. Though not vita, it is certnly helpful to an Meehan s March 12, 1976, testimony doe not have the weight of the actual formal offer itslf. Accordingly we have included CX 7013 in the reord.

. . . . . . 464 Opinion Fourlh: If an independent boy shop is unable to buy crash part from anyone of the GM dealers in its vicinity at the dealer price, GM would agee to sell these part to the independent dealer at dealer net price directly from its field warehous. (ALJX 17 Meehan testimony p. 122) This "safety valve" approach to IBS buying from GM warehouses is very close to GM's First Proposal"' Although it is too late to test the degree of good faith lying behind this Third Proposal (and it smacks not a little of a last minute grandstand ploy), nevertheless it does show that GM could slip out from under NADA's thumb whenever its purposes suited it. (b) The Miniwarehouse Plan Complaint counsel argue that at least one of GM's settlement proposals should have contained an offer to restructure the crash part distribution system by adoption of a "miniwarehouse plan. They argue that GM had such a plan on the drawing board, perhaps even in limited test operation, but aborted the program, at dealer insistence. (CAB 36) (22) The first problem with this theory is that there is no definitive explanation of the miniwarehouse plan" in the record. Three dealer witnesses (none of whom was ever a GM employee) testified as to their general understanding of its elements. (Stillman Tr. 8060 8071-72; Pohanka Tr. 4692, 4695-96; McCarthy Tr. 3490-91) From their often conflicting accounts it appears that the "plan" envisioned General Motors Part Division (GMPD) setting up franchises for the purpose of wholesaling crash part. The franchisees could be, but were not limited to, GM dealers. The mini warehouses were to supplement, not replace, GMPD' s then twenty-seven (possibly thirtysix) parts warehouses. It represented a limited opening up of the crash parts distribution system at the warehouse level. The record, which is extremely skimpy on the whole. issue shows that this "plan" was never significantly operational and although GM management may have been dimly aware of the concept, it did not endorse miniwarehousing in either theory or fact; dealers, to the extent they were even aware of the plan, considered it .. In the Send Propo GM alate that its first offersell to 188 had ben intended to apply "only in thos inatance in which the independent operator Waf unable to negotiate what he oorlidere to be a fair price. (TJhe warehous would colltitute an alternative BOurce, a lIety valve. (AL 13D) However, the Firnt Prpo, both 89 prente to the FI and 89 describe to dealers by GM, nowhere make! such 8. limitation. (AWX llg, ex 701OE, AL 16A, ex 7321) ao Complaint COUnBl's tetimonial ca rets entirely on four hostile witness, each a GM dealer or NADA offcial during the time of events they tetifed to. Havi dar thi much, their failure to cldl GMPD's General Manager Lewi Kauah-an equally hostile witness, but one infrnitely more know!edeable on the subjec weakens an alredy weak ca. With no reliable documents showig that the miniwarehous pltm bad any offcial ststua at GM, we are forc to rely upon the testimony ofthes dealern, de!pite its obviousoolf-orving nature. Opinion 99 F.

no threat, and forbore from lobbying GM concerning it. There is simply no boycott case arising out of the miniwarehouse scenario. The origins of the miniwarehouse program are obscure. Complaint counsel root it in the early 1970' , by virtue of an experimental dealer warehousing program in Phoenix, Arizona. (CPF 285)'0 However, the record contains only the barest (23Jrumor about this experiment as the only witness called to testify about it had no personal knowledge on the subject. He could give no details about Phoenix set-up, dates of operation (even to the nearest year) or success, if any. (Stilman Tr. 8070-72, 8142) Even allowing for the witness' bias towards GM we cannot see in his testimony the picture which complaint counsel paints of a viable, operational miniwarehouse program.

Dealers first became systemically aware of miniwarehousing in the second half of 1974, when the idea was floated before them by Lewis Kalush, General Manager of GMPD, an offce which does not carry the authority to change GM's distribution system. (Meehan Tr. 2227-29) Four witnesses either attended or heard of a meeting or meetings where the concept was discussed. However these meetings were not called solely or even chiefly to discuss miniwarehousing. GMPD was then in its second year as a separate division of GM. (IDF 162) The witnesses characterized the gathering(s) as getacquainted sessions organized by Kalush to air dealer complaints about the GMPD distribution system and discuss generally alternatives for improvement. (Faulkner Tr. 11803-11804) Significantly, the miniwarehouse alternative was not presented as a firm plan.

This was a concept that they had. They weren t saying whether they (were) going to establish it or not. This was a concept, a d:iussion that was going on about part distribution. (Stilman Tr. 8069) (24) The dealers perceived it as a pet project of Lewis Kalush, lacking in support from GM management" and of no threat to their exclusive .0 Complaint counsel P.&o rely on ex 7217, a document which from internal evidence was wrtten between 1972 and 1976, to give the miniwarehoUB prown historical authority. However, the document appeal" to be the product of GM'B Servce Seion (CX 7217 A), a GM unit which ha nothing to do with the ditrihution, warehousing or selling ofGMCP. (Meehan Tr. 2191, 2225) No witne1 Wil caned to identify the author of the document or Bhow that proam advance in ex 7217 Wil either conaidere or adopted by GM management. U Walter Stilman, a Buick dealer, Wil a fonner chairman of the GM Line Group of NADA'B Industrial Relationa Committe.

II It is unclear whether each witness refer" to the lie meeting, or whether &Orne are reling diferent Dries of a series of meeting. Witm- Stilma implies the latter (T. 8(), but his t.timony is the only BUpport for complaint counal'B asrtion that the miniwarhoUl pla Wil propo several times to grupe of dealer" between mid-1974 and ealy 1976. (CPF 28; CAB 36) .. There is &Ome conflct over whether GM-- oppo to GMPD-ffcel" attended the(se) first miniware hOUB meeting). Witness Faulkner rely GM president Cole being at the Augt, 1974 meeting (T. 11792); the thre other witnl! mentioned only GMPD offcel", mOBt notably KP.ush. (Stilman Tr. 801; Pohaka Tr. 4692; McChy Tr. 340091) 464 Opinion wholesaling rights in crash parts. This was made abundantly clear by NADA offcials' reaction to an unexpected pitch for miniwarehousing by Kalush at a meeting with select NADA offcials on April , 1975." The meeting was intended to be a briefing on the FTC investigation, but Kalush saw an opportunity to promote his project. As Frank McCarthy," Executive Vice President of NADA recalled: It was my reasonably clear recollection that even at the outst-but it became clear beause of the nature of the discussion-that the comments of Mr. Kalush were his comments as an individual, and did not reflect the opinion of even other members of the GM part division. (Tr. 3581) (25)1 mean, it was just very clear that Mr. Kalush, beause of the part distribution system, the way it is in General Motors, wanted the mini-warehouse concept, dealer stoking concept, implemented, beause that would come under the General Motors (Plant (DJiviion and it would come under him. And as part of sellng that idea to dealers, this is my clear impression, that he was trying to sell dealers on the miniwarehouse proposal as a solution to the crash part problems. (Tr. 3582) Complaint counsel assert, without any citations to the record, that the April 2, 1975, meeting crystalized NADA's opposition to the miniwarehouse plan. However, McCarthy s testimony shows neither a hostile nor defensive dealer reacton to Kalush's plan. Though the consensus was that more dealers would be made unhappy than happy by the plan (McCarthy Tr. 3492), NADA never considered the program a real enough threat to warrant pollng its membership for reactions. (McCarthy Tr. 3495) Again, allowance must be made for the witness' obvious bias in favor of respondent, but the record does not contain any testimony or documents'. which refute Mr. McCarthy. (26) The final time of significance in this history is May 1975. GM' .. The 1974 meeting(sJ donot oom to lwve involved any deael' holding offce in NADA. (Faulkner Tr. 11798- 99) By contrlit, the April 2, 1975, meeting woo initiate and attended only by NADA offcel': Jack Pohaka, President of the Industry Relations Committe; Gerge Erw, P"ibly Channan of the Servce and Part Committe; Walter Stilman, Chan of the GM Line Group; Paul Hei", Director of Rearch and Dealer Oprations; and Frank McCarthy, Executive Vice President of NADA. Represnting GM were Leis Kalllh and Jim Melica, Attorney in Charg of Trade Relation. (McCrthy, Tr. 34, 351) .. Neither Stilman nor Faulkner testified to this meeting; Pohaka recled it only in the IDQtt general outline. (Pohanka Tr. 469697) Thus McChy s testimony is the only detailed data in the record on thi subjec. .. Complaint counsel's sole support for the statement that the day afr April 2, 1975, NADA offcial held a meeting to oppo the miniwarehoWI progam iB ex 734A-F. We ca only sunniB that the document was aditte in order to give us the pleaure of reacquainting ourslves with that bae of law school evidence exainations, of triple heamay. The document consists of hadwritten note of Le Beaudry, a member of the NADA Part Committe, which happened to be meeting in NADA headquartm on April 3, 1975, in order to revi the part operations manua. (Budry Tr. 2750) If Beaudry s note are to be believed, Pohanka stopped by and told the Committe what KaIUBh and Melica had told the five NADA executives the day before. Neither Beaudry nor Pohaka haa any independent rellecion of the Apri3, 1975, meeting. To sdmit thc8 note for the truth of what Pohaka state was NADA's position, or, stil wors, what Wil GM' position with regard to any subjec goe beyond even the wide latitude accrded hearsy under the Commision Rule of Practice Seion 43().

Morever, even if the document ha any evidentiary weight it doe not prove what complaint counsel state. (Cn.tinue Opinion 99 F.

President Estes and four other offcials (none from GMPD) met with NADA' s President of the Industrial Relations Committee, Pohanka and Chairman of the GM Line Group, Stilman, to discuss "several items that GM dealers were interested in." (Pohanka Tr. 4691) Complaint counsel claim a major purpose of the meeting was for NADA to outline its objections to the miniwarehouse plan. However it is clear that many other topics were to be covered at that meeting. (Stilman Tr. 8143-4) Moreover, although as we have noted NADA was opposed to a totally open crash parts distribution system, it had not yet taken a position against the less sweeping mini-warehouse proposal. (Stilman Tr. 8145) Nevertheless one of fourteen items on the pre-meeting agenda presented to Estes was "Mini Warehouses . (CX 7205C) The record does not indicate whether Estes had knowledge in detail of the subject but he managed to convince the NADA offcials of his disinterest:

I remember using the word "mini-warehouse " and was surprised that Mr. Estes said what' s a miniwarehouse?"

We had to tell him.

Q. You mentioned that Mr. Estes was unaware of that term. After you informed him of the term, was there a disussion of miniwarehouses? A. As I recall, it was a completely new idea to him. I didn t see any point in pursuing it any further.

Q. So you just laid it on the table and left? A. AB I recall. (Pohanka Tr. 4694, 4698) (27) After this date the record is silent on dealer opposition to the miniwarehouse program.

Given this history, it is hardly surprising that GM's July 11, 1975 settlement proposal to the FTC made no mention of the miniwarehouse plan. At its strongest the plan appears merely to have been the pet project of a division head, who sought dealer aid in promoting it to GM management as he was powerless to implement it without endorsement by GM' s executive committee. (Meehan Tr. 2227-29) To the extent that the plan was tried out in Phoenix, it was dropped for causes other than dealer pressure. The record does not Rather, it confirm that the mini-warehous plan was perceived B8 Kalueh's e.one (CX 734B), that some deters liked the idea (CX 734) and is silent on the matter ofNADA oppoition to mini-warehousing. n The sin!;le line on ex 720 listin "Mini Warehous as a topic of diueion is the only documentary proof linkig the plan with uppe echelon GM management. The document WIl entered in the reord without benefi of explanatory testimony from either itb author, GM Vice Pmident Worden or its reipient, GM President Este. The document on itb face is merely a list of topiCl NADA wihe to disUS; it doe not indicate that GM manogemen had any independent information on thes topiCl. Thue even if GM Prident Cole attepded a meeting where minwarehousin was dius in 1974 (se n. 33 8l.pm) is entirely poible that offcia awarenBS of the concept did not over to Prident Eete' adminisration. ca 464 Opinion show that the mini-warehouse plan was either operational on July , 1975, or being seriously considered by GM at that time. Nor does the record show that NADA launched any specific lobbying effort against the mini-warehouse proposal.

(c) Legal Analysis Whether known as boycotts or concertd refusals to deal collective effort to exclude a party from the marketplace are ilegal per se under Section 1 of the Sherman Act. Northern (2S)pacific Ry. Co. v. United States 356 U.S. 1 (1958); Klors, Inc. v. Broadway-Hale Stores, 359 U.s. 207 (1959). Even the Supreme Court, however, has recently acknowledged that the decisions on what constitutes necessary elements of a boycott in violation of the Sherman Act reflect a marked lack of uniformity in defining the term. St. Paul Fire Marine Ins. Co. v. Barry, 438 U.S. 531, 543 (1978). Although the finding of a boycott or concerted refusal to deal invariably turns on the facts in the cases, no conclusive fact-pattern has emerged which spells out precisely what a boycott is. As a consequence the cases in this area are often confused. Grouping them under the categories of primary and secondary boycott'. provides an analytical framework which helps to identify useful precedent for the matter at hand.

The first type of case deals with a "primary" boycott, where a number of economic actors at one level of the productive or distributive process either discontinue economic relations with an actor or actors at another level, or predicate continuance of economic relations only on certain terms. Paramount Famous Laky Corp. v. United States, 282 U.S. 30 (1930). A distinguishing feature of a primary boycott is that the resultant economic harm is suffered businesses which are not competitors of the members of the combination. The case before us is not a primary boycott situation .. Scholat of thia issue have debate whether the draconia forelosure of inquiry which is the halmak of per 1If! analysis ehould be applied to every concert refer to deal In particular, their concern is that market effec and speific purp should be weighed when the group exercising concert power is noncommen:ia; or when the mai purp and effec of the boycott tire not proteion of the COnHpiraool'' profits. , e.g., Bird Shennan Act LimitatioIl on Noncommercia Concert Refus to Deal, 1970 Duke L.J. 247; Baber, Refusas to De Under the Antitrwt Laws 103 U. Pa. L. Rev. 847 (1955); L. Sullivan, Hadbok of the Law of AntitTUt 259 (I977). To clarfy when the us ofper seanalyais is a.pproprite, some commentatorn have propo mutually exclusive definitions of the term "boyctt"and "concert refua to dea". Sullvan supra 25; Note, Boycott: A caSpeifcbeforeDeflDition11 allegesLimitsa theclalicApplicablityexercis ofofconcertthe Per SepowerRule,by71tra.dernNorthwesternat ones.levels. Rev.of the818distribution(1977). However,pr0 the prote themslves from competition or potential competition at that level. Therefore, we wil us the term boyctt" and "concert refua to dea" interchageably; we expre no opinion lI to the merit of any of the IimitingdeflDitiolU.

n The term ar derive from Bird, Shennan Act LimitatiolU on Noncommercia Refus to Deal, 1970 Duke J. 247 (1970). Prfesr Bir alo describe a third variety ofconcerl refuals to dea where a group establishes a joint facility or a trade WIiation and limits aco to it. Th situation is far removed from the ca before and therefore nee not be conaidere here.

Opinion 99 F.

since GM and its dealers do not occupy the same level of the productive process, and the victim group (IBS) is competitive with some of the members (dealers) of the alleged combination. In a secondary boycott, a group threatens an economic actor or actors at another level of the productive or distributive process to force them to refuse to deal with someone else-usually a competitor of the boycotting group. A distinguishing feature of a secondary boycott is that the resultant economic harm is directly caused by a conscripted neutral, not the boycotters. (29) Secondary boycotts may be further divided into two subgroups depending upon whether the coerced neutral stands above or below the boycotters in the distribution chain. United States v. Parke Davis and Co. 362 U.s. 29 (1960), is an example of a Sherman Act Section 1 violation found in coercion flowing down the distribution chain. There a pharmaceutical company refused to deal with wholesalers in order to elicit their wilingness to deny its products to retailers and thereby help force the retailers' adherence to its suggested minimum retail prices. Such cases are not analogous to our situation, where the alleged pressure rose from the distributor level to the manufacturer.

We must look, then, for precedent in cases dealing with pressure by entities at a lower level of the distribution chain upon their supplier.

Turning to such cases, we find none where coercion was found solely on the basis of such a blustering, but ultimately toothless course of conduct as the GM dealers engaged in during 1974-1976. NADA never once (on this record at least) threatened GM with the one sanction which would be significant to it: loss of new car sales. Cf U.S. v. General Motors 384 U.S. 127 (1966). As the ALJ noted, our situation is a long way from the naked wielding of buying power in Hershey Chocolate Corp. 28 F. C. 1057 (1939), affd 121 F.2d 968 (3rd Cir. 1941).

However, complaint counsel are also correct when they argue that direct proof of the wielding of such power is not a necessary element of a violation of Section 1 of the Sherman Act. (CAB 37) There have indeed been cases where a response to umere complaints" was held to be an act under compulsion. What complaint counsel miss however, is the fact that in these cases the courts, finding the complaint did not constitute an obvious threat, inferred the existence of the iron hand inside the velvet glove from the fact that the target suddenly and completely capitulated to the boycotter s wil. These cases also indicate that the more irrational the changed , 464 Opinion business pattern, from the viewpoint of the target, the more suspect is its motivation.

In this matter, GM never committed itself to do what the dealers wished. GM's constant brush-off of its dealers' complaints contrasts sharply with the situation in Ford Motor Co. v. Webster Auto Sales, 361 F.2d 874, 877 (1st Cir. 1966), where, in response (30 Jto a dealer complaints, Ford sent letters to its dealers requesting them not to bid on "company cars" for the purpose of wholesaling them. Another contrast may be found in Bowen v. New York News Inc. 522 F. 1242 (2d Cir. 1975), cert. denied 425 U.S. 936 (1976), where, in response to franchise dealer complaints the newspaper publisher harassed and terminated supply of its papers to franchisees who sold those papers to independent dealers in competition with the franchised dealers; and in u.s. v. General Motors Corp. 384 U.S. 127 (1966), where within two months of receiving complaints about dealers who had business dealings with discounters, GM elicited from each such dealer a promise to discontinue the practice, and set up a system to police compliance with the agreements. Complaint counsel argue hotly, and with some logic, that since what the dealers requested of GM was inaction, compliance with dealer "requests" cannot possibly be shown by any affrmative response-all GM had to do to acquiesce was follow the course it had been following already and keep its crash parts distribution system closed to all but franchised dealers. However, this argument infers too much. Even in cases where direct action is taken by the supplier and such action is precisely what is requested by complaining distributors. the inference of concerted action a conscious commitment to a common scheme " is not made automatically. Edward J. Sweeney Sons, Inc. v. Texaco, Inc. 637 F.2d 105 (3rd Cir. 1980), cert. denied 451 U. S. 911 (1981). In Sweeney, the defendant gasoline supplier could show that it had its own reasons for terminating a dealer, and did not object to his pricecutting practices which had caused competing Texaco retailers to complain. Consequently, there was no concerted action, even though the result was the same as if the threat had been heeded.

When no change to a pattern of business conduct is the gravaman of the complaint, it is particularly diffcult to overcome the inference that the manufacturer s choice not to move is based on those same unilateral business reasons which led it to adopt the system in the first place.'o Thus, in Aviation Specialties Inc. v. United Technolo- .0 A manufacturer s goo faith choice of a distribution system may stil be an arbitrary one. A later section of thi opinion distl the objective rea.nableoes of GM'B commitment to a selective distribution system for crBBh part; at this time, however, we Ilre concerned only with divining whether that decision was unilateral or not. Opinion 99 F.

gies Corp. 568 F.2d 1186 (5th Cir. 1978) a would-be distributor of repair parts for the Pratt-Whitney P-T6 (31)airplane turbine engine challenged Pratt-Witney s refusal to deal with him as a violation of Section 1 of the Sherman Act. In affrming summary judgment for the defendants, in part because plaintiff failed to prove a conspiracy between Pratt-Witney and its distributors, the court stressed: ASI bears a particularly heavy burden beause Pratt-Witney set up its distribution system in 1964, long before ASI began operations, and the structure of the system has not changed perceptibly since its inception. 568 F.2d at 1192. In Aviation Specialities, the plaintiff had absolutely no evidence of dealer pressure on the supplier to keep the distribution system closed, whereas in this case we have a history of dealer efforts to influence the supplier. Nonetheless, the burden of showing collusive conduct is stil particularly heavy when the challenged action is a decision to maintain a long-established distribution system. (d) Conclusion Here, by looking at the total course of conduct rather than merely the final moment, we are persuaded that GM acted in its own selfinterest, rather than at dealer behest. The Tighe notes and similarity of the February proposals raise a question of conspiracy. But against these must be set the entire pattern of conduct between GM and NADA during 1975 through March 1976. That pattern reveals that , like any manufacturer, preferred to have its dealers' good will. Accordingly, when it cost GM nothing to placate the dealers it did so: giving general expressions of sympathy during various 1975 meetings and changing the First Proposal after the FTC showed no sign of accepting it. On the other hand, GM was quite ready to ignore the dealers when there was advantage in doing so. Hence, the First and Third Proposals offered without consulting the dealers; the constant refusal, during those 1975 meetings, by GM to commit itself to any hard and fast position. Moreover, NADA, in our opinion, read the pattern the same way. Throughout 1975 dealers were in uncertainty and despair over GM's intentions towards them. They did not threaten GM with economic reprisals. Instead they argued and pled and tried to enlist alles in Congress and the FTC. These are not the actions of successful boycotters. (32) Abuse of Monopoly Power Through Leveraging Complaint counsel advance several theories under which GM' distribution system is an abuse of monopoly power. Two of these theories, uleveraging" monopoly power and "extension" of monopoly 464 Opinion power, are virtually identical and not applicable to this case due to the same fact which caused the horizontal boycott theory to founder: GM' s minuscule presence in the dealership level of the crash part distribution chain.

The court have long held that it is an abuse of monopoly power for a monopolist to use its monopoly power in one market to extend or leverage that power into another market. Otter Tail Power Co. v. United States 410 U.S. 366 (1973); United States v. Griffith 334 U. 100 (1948); Berkey Photo Inc. v. Eastman Kodak Co. 603 F.2d 263 (2d Cir. 1979), cert. denied 444 U.S. 1093 (1980); Sargent- Welch Scientific Co. v. Ventron Corp. 567 F.2d 701 (7th Cir. 1977), cert. denied, 439 S. 822 (1978); United States v. Aluminum Co. of America 148 F. 416 (2d Cir. 1945); United States v. Klearflax Linen Looms, Inc. 63 F. Supp. 32 (D. Minn. 1945). However, the majority of these cases arose from situations which suportd a finding of attempt to monopolize as well as abuse of monopoly power by leveraging activities. They therefore have limited guidance to analysis of the facts in this case, where attempt to monopolize was never an element of the case and cannot be proven. Two closely intertwined requisites for an attempted offense are totally lacking: (1) specific intent to control prices or destroy competition United States v. Aluminum Co. of America, 148 2d 416 (2d Cir. 1945), and (2) dangerous probabilty of success United States v. Swift Co. 196 U.S. 375 (1905). To ilustrate this we need do no more than refer to facts already summarized concerning GM' s historical presence at the dealership level through its MHD financial arrangements.

As noted, General Motors' only presence at the retail and installation level of crash parts distribution is its financial interest in the MHD dealerships. Even assuming this financial interest amounted to functional control, a market presence which (33)has not risen above 3% in I3 years does not remotely suggest that GM is moving progressively closer to monopoly power at the dealership level. 41 On the question of intent, again we note that the record is devoid of evidence that General Motors' crash part distribution policies are ., By contraat, in Otler Tail the defendant wholesaler of electrical power ha achieved a 91% share of the leveraged market for reta power. 410 U.S. at :370. InGriffith the w.ngeroUl probabilty stadar waa not llpeificaly dill, as the CI was remanded for prof of effec. However, the court note that defendlUt fim exhibitors had increas their share of single-theatre toWI from 51 % to 62% over a five yem period. 334 U.s. at 102. It was during this period that defendants exercis their poled buyi power to Cl1ure that members of the circuitgol exc\Ulive rights to first ru fIIl, to the detrient of their competitors. ld. at 104. Tially, Alcin the defendant UB its monopoly power in aluminum ingot to imp! a price squee 00 purchllTI which compete with it in the mlUufacture of aluminum sheet; by thi melUs Alcoa, already the largelt maker of aluminum sheet, elim.inate halfthc compaies competing-with it in that m.arket. 148 F.2d at 436. Opinion 99 F.

established or maintained in order to benefit the MHD dealerships. Nor is the mere choice of a selective distribution system, even by a manufacturer who is also a distributor, in itself the sort of invidious conduct which support an inference of intent to prevail over competitors by improper means" (34) Nevertheless we must stil determine if General Motors' conduct fits within that line of cases which holds that leveraging monopoly power can violate the antitrust laws even when it does not amount to an attempt to monopolize. Complaint counsel rely heavily upon Berkey Photo Inc. v. Eastman Kodak Co. 603 F.2d 263 (2d Cir. 1979), cert. denied 444 U. S. 1093 (1980) and Sargent- Welch Scientific Co. v. Ventron Corp. 567 F. 2d 701 (7th Cir. 1977), cert. denied 439 U.S. 822 (1978).

Both cases differ from this one in terms of mode, impact and especially purpose of the leverage. In the matter of mode the difference is least and the analogy strongest. While neither Berkey nor Sargent- Welch involved a refusal to deal, each did concern a mode of leverage which was not pernicious on its face, but instead could have been a reasonable business act. In Berkey, defendant Kodak introduced new products; in Sargent- Welch a manufacturer terminated a dealer, ostensibly because of the dealer s failure to represent the manufacturer adequately.

In neither Berkey or Sargent- Welch did the monopolist have as small a share of the leveraged market as GM does in the crash parts installation (dealer bodyshop) market. Moreover, in both cases the leverage affected a new segment of a market, such that the resultant distortion in favor of defendant might be expected to grow. By contrast, GM' s share of the bodyshop market has remained static for a long time. Thus impact on competition was potentially more severe in Berkey and Sargent- Welch than here. This fact is closely alled to the purpose element of these cases. Although the specific attempt to monopolize standard has been diluted to a general intent to gain a competitive advantage in the downstream market, even this lesser intent cannot be found in General Motors' choice of a distribution system. It would be an .. By contraat, inKkarf the defendant ru manufacturer and distributor tok aeveroJ clearly preatory stepa auch B8 refu.ing to fll orders for a rivw ditributor, Floor Product, Bfter learning that Floor Products hoo underbid it for B government contract, and aakig itb distributors not to undercut it by aelling to Floor Products. Morever, KJearfax s Generw Manager speifically announce "My plan is definitely to Bquee Floor Producta outofthigovemmentbUlineo... . " 63 F. Supp. at 36. .. As we note in I. DuPon! ck Nemourt & Co, 96 F. C. 65 (1980), the deg of market power which indicate a daerou. probability of BUcc may VBry with the nature of the chalenged conduct-Id. at 725 Here, however, GM'a choice of a aeJecive distribution a)'tem is so innocuous in nature that the question of applying the aliding Bele doe not ari.

.. It WB6 never actu811ymeasured in either Berkey or Sargent- Weich-both C8 were remanded for further fmdin on exactly this issue and actUed out of court before findings could be taken. 464 Opinion example of the tail wagging the dog to infer that GM chose to distribute crash part selectively in order to protect its MHD dealerships' bodyshops from the competition of the IBS. (35) C. Abuse of Monopoly Power Through Refusal To Deal Legal Analysis Complaint counsel are on firm ground at last when they turn to the line of cases which concern abuse of monopoly power over a scarce resource which other firms are under a commercial compulsion to use. Otter Tail Power Co. v. United States 410 U.S. 366 (1973); Silver v. New York Stock Exchange, 373 U.s. 341 (1963); Associated Press v. United States, 326 U. S. 1 (1945); United States v. Terminal R. Assn. 224 U.S. 383 (1912); Fulton v. Hecht 580 F.2d 1243 (5th Cir. 1978); Hecht v. Pro-Football, Inc. 570 F.2d 982 (D.C. Cir. 1977), cert. denied 436 U.s. 956 (1978); Gamco Inc. v. Providence Fruit & Product Bldg. 194 F.2d 484 (1st Cir. 1952), cert. denied, 334 U.S. 817 (1952); Lake Carriers Assn. v. United States 399 F. Supp. 386 (N. Ohio 1975); Grand Caillou Packing Co. 65 F. C. 799 aff'd sub nom. La Peyre v. 366 F.2d 117 (5th Cir. 1966)." In each of these cases FT monopolistic control over a unique or essential product or facility caused significant competitive harm when coupled with a refusal to deal with a portion of that class of persons who were under a commercial compulsion to use the product. In other words, the monopolist' s refusal to deal was judged ilegal because it created a haves " VB. "have nats" situation in the downstream market. In some of the cases the monopolist itself, in its secondary distribution function, was in the downstream !Chaves" group, and so benefited from the selective refusal to deal. Thus, some of these cases appear to fit under the rubric of "leverage" where, as we have seen an abuse of monopoly power is found when the facts show that the monopolist intended to bolster its downstream market position by use of its upstream monopoly power. However, the rationale of these essential product cases" goes beyond that (36)of the leverage cases. In essential product cases the abuse of monopoly power lies in the failure to make a scarce resource available to all potential users on nondiscriminatory terms, not merely in any incidental benefit to the monopolist' s position in the secondary or downstream market. The court' focus in such cases has been on the unlawful harm to .. Als part of this lineof C8 is the Commision s decilion inthe Reuben H. Donnelle Corp. 95 F. C. 1 920 (2d Cir. 1980),cert. denumreu d sub /wm. Fedrol Tr Commision v. Om w.l Airline Guid, lru. 63 F.2d 101 S. Ct. 1362 (1981). The Commision s position is that the Send Circuit'a reversl ofDcmnelley WIl erroneous. Neverthele!, we do not rely upon theDonnelle decision in this ca; but until and unless8 it is repudiate by the Supreme Court we hold to our interpretation of the ca law on arbitrary reflU18 to deal by monopolists, which ha ben eapoua by the Fifth Circuit in La Pere, supra GENERAL MOTORS CORP. 581 Opinion 99 F.

competition, not the gain to the monopolist. Similarly, the fact that in some of the cas!JS the competitive injury was caused by a joint refusal to deal does not weaken their precedential value to this case. It has long been established that "the existence of a combination " is not an "indispensible ingredient" of an unfair method of competition under Section 5 of the FTC Act. FT v. Cement Institute 333 U. 683 (1948). The Commission, proceeding under Section 5, may properly draw upon the policies expressed in the Sh!Jrman Act' prohibition of joint nifusals to deal. Those policies mirror the conC!Jrns of the scarce resources cases: monopolistic power should not be used to discriminate among competitors in an adjacent market, if that discrimination is arbitrary and causes substantial competitive injury. Again, the emphasis is primarily upon the competitiv!J dislocations caused by the discrimination, and only secondarily upon the b!Jnefits accruing to the discriminating party or parties. The earliest of these cases makes this abundantly clear. Terminal , supra concerned a group of railroad companies that controlled all rail access to St. Louis by virtue of owning all the bridges into that city. There was no showing that the monopolist association had used this power in any way against rival railroads. The Supreme Court fashioned the bottleneck theory out of concern for potential abuse of monopoly power by a future denial of access to thj essential facility. The Court's concern was not that the Association would contribute to its member railroad companies' power, but rather that non-memoor railroad companies would suffer from lack of access to the facilty.

We fail to find in either of the contracts referred to any provision abrogatig the requirement of unanimous consent to the admision of other companies to. the ownership of the Terminal Company, through counsel say that no such company will now find itslf excluded from joint us or ownership upon application. That other companies are permitted to us the facilities of the Terminal Company upon paying the same charges paid by the proprietary companies sems to be conceded. But there is no provision by which any such privilege is accorded. (37)It cannot be controvert that, in ordinary circumstaces, a number of independent companies might combine for the purpo of controllng or acquiring terminals for their common but exclusive us. In such cas other companies might be admitted upon terms or excluded altogether. If such terms were too onerous, there would ordinarily remai the right and power. to construct their own terminals. But the situation at St. Louis is most extraordinary, and we bas our conclusion in thi case in a large measure, upon that fact. The "physical or topographical condition peculiar to the locality," which is advanced as a prime justification for a unified system of terminals, constitutes a most obvious reason why such a unified system is an obstacle a hinderance and a restriction upon interstate commerce, unless it is the impartial 464 Opinion agent of all who, owing to conditions are under such compulsion as here exits touse its facilities. 224 U$; at 40 405.

It requires a close reading of Gamco, supra, to discover that the corporate owners of defendant Providence Building were actually wholesaler competitors of the. plaintiff. The Court' s emphasis is not on the joint activity, but rather on the essential nature of the fa.cilty and the substantial harm suffered by Gamco in being denied access to it.

But it is only. at the Building itslf that the purchasrs to whom. a competing wholesaler must sell and' the rail facilities which constitute the most economica method. of bulk transporttion are brought together. To impose upon plaintiff the additional expenses of developing another site, ' attracting buyers, and tranhipping hi fruit and produce by truck is dearly to extract a monopolist's advantage. 194. at 487.

In the La Peyre case, the La Peyre family enjoyed a lawful monopoly in certin machinery used in shrimp canning; this machinery was so effcient that its use became essential to compete in the shrimp canning business. The Commission challenged the family s practice of leasing its machines at twice the rental rate in the Pacific Northwest as in the Gulf Coast, and found that the discriminatory excess rental charge was the cause (38Jof many West Coast wholesalers and packers operating at a loss, sometimes to the point of being driven out of business. Grand Cailou Packing Co., 65 C. at 841-845. The appellate court emphasized that this practice did not involve Robinson-Patman type discrimination but concerned a much broader, more far-reaching issue: "the duty of a lawful monopolist to conduct its business in such a way as to avoid inflcting competitive injury on a class of customers." 366 F.2d at 120. The court clearly held that the leasing practice of a single company constituted an unfair method of competition because it involved "the utiiztion of monopoly power in one market resulting in discrimination and the curtailment of competition in another. Id. at 12I. Respondents in the case before us mentioned that the La Peyre family had shrimp canning operations on the Gulf Coast; thus their motive for leasing machinery at discriminatory rates could have been to cripple growing competition from the West Coast. However, the decision of the Fifth Circuit rests on broader grounds, as the above quotation demonstrates. Moreover, the Fifth Circuit subsequently explained its own decision in La Peyre as holding that the exercise of monopoly power to injure competition in an adjacent market itself violates Section 5. Fulton v. Hecht 580 F.2d at 1249 n. The concern in these cases that competition will be harmed by the GENERAL. MOTRSCORP. 583 Opinion 99 F.

refusal to deal with a class of competitors is very strong. According' ly, the product or servce need not be utterlyindispensible Hecht, 580 Fc2d at 992, or completely incapable of substitution Gamco, 194 2d at 992; and a Sherman Act violation has been found even where customers are not totally excluded from the market, but merely placed at a competitive disadvantage. Associated Press 326 U .s. 13. Where an essential product or facility is involved, courts may even require procedural and substantive guarantees of fairness to justify exclusion or discriminatory treatment of cust(mers. Silver 373 U.S. at 361; Gamco 194 F.2d at 487.(39) The foregoing discussion of the significance of harm to competition in the line of precedent does not mean that a monopolist has an absolute duty to deal whenever some such harm is shown. On the contrary the demonstration of substantial injury to . competition is merely the trigger which sets off a rule of reason analysis of the monopolist' s reasons for refusing to do business with all. Because of the nature of this inquiry the duty to deal arises rarely, and only after an exacting balance of the equities. A supplier s general right to choose its customers, enunciated in United States v. Colgate Co. 250 U.S. 300 (1919),4' wil not be questioned unless and until the harm shown is substantial, and affects existing competition. When that degree of harm is established, the monopolist must show that its decisions which cause the harm to competition were made for substantial business reasons, not arbitrarily. And even if the decisions were made arbitrarily, the Commission wil not impose a duty to deal if the order would require the Commission or enforcing court to assume a continuing role in supervsing business discretion. (40) Monopoly Power Whether GM is a monopolist is the threshold question under this .. The seller's freom to trade enunciate in Colgte doe hsve limits. It i6 only a general right, "neither ablute nor exempt from relation:' Loroin .JQurr1 Co. v. U.S. 342 U.S. 143, 155 (1951) The circumstace under which a company s refu8J to deal is not proi: by due general right have ben and stil are being edu.lUBtively explore on a cahy-c basis sinre Colgate. In the monopoly context it is well to remember thatColgte in no way repudiate the esntial facility docrine enunciate in Terminal RailrO decided seven yearn earlier. The market afec by Colgate's alleged practice was the manufacture of sop and toilet articles, a market tol.l1y dierent from that inTerminal Railro. The n.traod tenninwWI an esntil faciliy; by contrast altenwtives tv Colgate's products were readily available. The importCe of thi diinction was lau.r ernphaiz in Unite State 11.Arnld &hwinn CO., 38 U.S. 36 (1967). There, the Supreme Court retrict a manufacturer s right ro corime hie 8aes to selected dealers ro situations where "competitive product are redily available ro others" Dr where "other and equivalentbranda. . . ar rediy available in the market." lri at 376.&eal8 Eldr-Bet1n Stam; Corp. 11. Fetrate DepartmentSkim;, Inc 459 F.2d 138 (6their. HJ72). Where an esntial product is involve, therefore, theColgte doce provides the monopolist no proteion.

464 Opinion theory. To answer it we must. determine the relevant product market." Because there are no close substitutes for new GM crash parts, and therefore GM has nearly unfettered pricing discretion, the ALJ found the market to be new GM crash parts for GM automobiles and light trucks. (IDF 33 392) We agree with this finding. In this market GM has a 100% share, since it is the sole producer of all new GM crash parts. (IDF 68) Respondent contends that the ALJ wrongly excluded salvage (used) parts as reasonably interchangeable substitutes for crash part.'. . Our review of the record convinces us that the ALJ was correct in his ana.lysis of the interplay between the two products: while there is limited interchangeability, it does not amount to any degree of effective competition. Salvage parts are very imperfect substitutes for new crash parts. Availabilty is extremely limited in the early years of any type of crash part. Moreover, salvage parts cannot be used as easily as new crash parts: the used parts generally have to be repaired and are often not as good a "fit" to the car. Prices of the two types of parts appear to be highly independent. The fact that limited substitution is observed does not disaffrm the existence of monopoly power. At monopoly or exclusionary prices, we would expect to see some substitution, if at all possible. We therefore endorse the ALJ' s classification of General Motors crash parts as a separate market on the basis of five of the Brown Shoe criteria: (1) specialized vendors; (2) peculiar characteristics and uses; (3) industry and public recognition; (4) distinct prices; and (5) lack of mutual price sensitivity. We incorporate by reference here rDFs 219-242 and 306-329. See also United States u. Aluminum Company of America 148 F.2d 416, 425 (2d Cir. 1945); Aunet Inc. u. Federal Trade Commission 511 F.2d 70 (7th Cir. 1975); United States u. CBS, Inc. 459 F. Supp. 832, 838-39 (C.D. Cal. 1978). (41) 3. Harm (a) Locus of Competition in Crash Parts The question of whether a monopolist has arbitrarily refused to deal cannot be answered without first determining the nature and extent of harm caused by that refusal to deal. The more grave the effects on competition, the more substantial must be the monopolist' s justification for its actions.

., It is clear that the relevant gooaphic nlfUket is the Unite State IU a. whole, (lDF 335) .. Respondent al rooms to arge that the proper market is new automobiles. (RA 28) It doe not cl'plain in any data why the (cnu h) pa is ineXtricable from the whole for purp of product tnElrket defmition. Certly the price of crWlh pa ha litte, if liy, effec on competition in the aae of 1'ew automobiles. Nor are GM crash pam inrercha.geable with crEWh part for otherroanufacturel'' cars. In this regaro we note that in FLM Collision Parts, supraboth the Ditrict Court and Send Circuit accepted that Ford ha a monopoly on cNlh part manufacture exclllively by Ford for Ford cars. 543 F.2d at 1030; 40 F.Supp at 227-28, 24. GENERAL MOTORS CORP.

Opinion 99 F.

Thus, our second step in constructing the balance is to find the locus of competition in crash parts. To do so requires a clear understanding of the functions performed with regard to crash part by the only entities currently involved as conduits in the intermediate (post-manufacturer, pre-retail) stage of crash parts distribution: GM and its franchise dealers. Below the manufacturing level GM has only one function, that of warehousing or primary wholesaler (through GMPD). The dealers are in a more complex situation. While most dealers install crash parts in their own bodyshops, only some dealers both wholesale at a secondary level and install a dealership crash part. Keeping a firm conceptual distinction between the wholesaling and installng functions is essential. For the remainder of the opinion we wil refer to dealer-installers and dealer-wholesalers in order to help clarify this vital functional distinction". All relevant business transactions with the IBS are performed by It is in this capacity, asdealers in their role as dealer-wholesalers.'o suppliers of crash part to the IBS, that dealers receive wholesale compensation from GM.

The second function a dealer may perform is to operate a dealer body shop where crash part are installed. It is in this bodyshop operation which, as an installer and retailer of crash parts, a dealer competes with the IBS in collsion repair work. Dealers do not receive wholesale compensation for crash parts which are used in their own body shops. (42) It is helpful to keep in mind that every individual crash part must be handled by a dealer before it is installed by either a dealerinstaller or by an IBS in a consumer s vehicle. However, crash parts which are installed by an IBS are only handled by dealers in their capacity as dealer-wholesalers. That is, any individual part purchased by an IBS from a dealer-wholesaler is never handled by a dealer in its capacity as dealer-installer. While this distinction may seem obvious and somewhat trivial, failure to make it resulted in the ALJ' s erroneous determination that the wholesale compensation isplan, rather than the selective crash parts distribution system, what gives dealer-installers a competitive edge over the IBS. This framework also makes it clear that wholesale compensation should not be subtracted from GM's warehouse price in computing the cost 4. We rerve the word "deaersp" fot ocions when distinction between distribution functions if l1necry.

00 AJ describe at IDF 76, deer-wholesalers roay al supply crash pam to cuswmera other than IB, an such tranctioll mayor may not qualify for wholege compensation under the WCP. Thes tranctol' have n beng upon the parent diu8itm.

464 Opinion to a dealer of crash part which wil be installed in his body shop, as oppoed to used by IBS.51 The WCP rebate is simply irrelevant to a comparison of the prices which IBS and dealer-installers pay on crash part to be used in their respl)ctive body shops. On the point of competitive injury, complaint counsel argue that two functionally . distinct groups of businesses are affected: the independent boy shops (IBS) and the independent wholesalers (IWs). We agee that both are adversely affected (though to different degrees) By GM' s refusal to deal, but only in the case of the IBS does the lost opportunity to buy crash part directly from GM also translate into harm to competition.

1. Indpendent Wholesalers IWs playa large role in the autOlnotive aftermarket. (See generally, Nelson Tr. 13719-20, 13758-13761; IDF 49, 129-140) They may be either warehouse distributors (redistributing wholesalers) or jobbers, who sell directly to repair outlets. For the purposes of this analysis it is not necessary to distinguish between jobbers and 52 The key factor is that neither currentlywarehouse distributors. handles crash part. General Motors is the only entity currently engaged in primary wholesaling 01 warehousing of crash part. (43) If GM sold crash part to IWs they would replace or rival GM' thirty-six GMPD warehouses and seven part plants. They are, to borrow a term from merger law, potential entrants" into the business of crash part wholesaling. There is no actual competition at this distribution level: GM by forward integration has pushed its monopoly over the production of crash part into the first level of crash parts. distribution.

Forward vertical integration by a manufacturer can be implemente for legitimate competitive reasons, such as distribution effciencies and profit maximizing based on lower prices and higher output. Accordingly, it is condemned under the antitrust laws only when it has the collateral purpose or effect of achieving some mticompetitive advantage for the integrating company. For exam- ,Ie, a manufacturer may not integrate forward if that action ,mounts to an attempt to monopolize the downstream distribution vel. , Eastman Kodak Co. v. Southern Photo Materials Co., 273 8. 359 (1927); and commentators have suggested that forward ., ThOB portiOll of IDFsll9-128, 401, 40 which state or imply otherw ar speifcay rejec. .. More proiB defmitionB ar given a.t Ab6n Tr. 12490; Nels Tr. 13719-20; tr alao Davin, The rketin of Automotive Par, Michian Buainef Stl1dies, Vol 12. No. 1 at 956958 (An Arbr: Univ8n1ity ,than 19&1).

'. We do not nee to relve the confct in the rerd over whether IWs have the desire as well as the wilty to enter int. crah pa wholes.

GENERAL MOTORS COlt!'.

Opinion 99 F.

vertical integration should not be allowed when it is a device for maintaining monopoly at the manufacturing level by raising entry barriers to would-be manufacturing competitors who could not independently distribute their goods effciently. See Areeda and Turner, III Antitrust Law U726d3, U726d5 (1980). However, as we have already acquitted GM of any anticompetitive purpose in its choice of a distribution system, the examples above are not relevant to our analysis.

We do not think the duty to deal requires a monopolist to set up rivals to itself; the duty merely requires the monopolist not to discriminate arbitrarily between existing classes of competitors. Similarly, the duty to deal does not require a monopolist to create competition in a subjacent market where none exists. Thus, the IWs position as mere potential competitors of dealer-wholesalers does not entitle them to the advantages of a place in GM's distribution system. Even a monopolist has a general right unilaterally to decide with whom it wil deal in the first instance, absent any improper purpose. United States v. Colgate Co. 250 U.S. 465 (1919). Certainly if failure to expand a distribution network of a fixed number of primary distributors were ilegal, every selective distribution system would be in jeopardy. (44JAviation Specialities, Inc. v. United Technologies Corp. 568 F.2d 1186, 1192 n. lO (5th Cir. 1978). We do not think that the duty to deal sweeps this broadly. Only when the monopolist's refusal to deal creates disequalities among existing competitors wil the monopolist have to justify its choice of selective distribution system.

2. Independent Body Shops IBS are actual, not potential competitors of dealer body shops in the retail and installation of crash part. Accordingly, we develop the harm side of the equation by analyzing the extent to which GM' refusal to deal retards the IBS' competitive strength. (b) The IBS' Competitive Position What the IRS Pay The record is uncompromising on the fact that the IBS pay significantly more for crash part than GM dealer-installers. General Motors sells all crash partwhether or not eligible for wholesale compensation-to its dealers for list less at least" 40%. (IDF 119 50 Approximately 50% of deaeTI' ordel" are routine sto orders (PAD ordel1) and Il are eligble for aJ additional15% disnt. This bri the pncdown to 43% off list. (lF 120, 122, 1901), 195, 24) "co FEDERAL TRADE COMMISSION DECISIONS 464 Opinion 248) In contrast to this dealer-installer cost" everyone of 26 IBS witnesses from seven trade areas testified that his invoice cost on crash parts during the last six-to-seven years averaged between 22% and 31 % off list price. (Neibling Tr. 2533, 2534, 2538; Craft Tr. 2910; Perschall Tr. 3086; Clouatre Tr. 3635; Trepagnier Tr. 3845; Lakatos Tr. 4008-009; Clark Tr. 4218; Serwacki Tr. 4811, 4840; Whitman Tr. 4991, 4992; Barney Tr. 5231; (45)Baker Tr. 6183, 6184; Hershey Tr. 6569; Weatherford Tr. 6712, 6715, 6716; Newman Tr. 5713, 5714; Crigger Tr. 5785; Brokaw Tr. 5977; Smith Tr. 7353, 7354; Rouse Tr. 6929 6930; Albertin Tr. 8241; Finkle Tr. 9324) Seven GM dealers and parts managers testified that their sellng price (wholesale price) to IBS was generally within the range of 22%-31 % off list. (Schaeffer Tr. 10665; Sutliff Tr. 11019--20; Tribo Tr. 10847-848; Bogard Tr. 10489-490; Mehaffey, Tr. 11222; Boyd Tr. 11853-54; Denton Tr. 12011--12) This testimony is confirmed by a stipulated summary of some 5 000 invoices" issued by 82 GM dealers to the IBS witnesses. The summary, reproduced below, shows average crash parts discounts for the seven trade areas over three years: AVERAGE DISCOUNT (% off list price) Trade Area 1974 1975 1976 Buffalo, N.

Mansfield, Ohio Cleveland, Ohio New Orleans, La.

St. Louis, Mo.

Spokane, Wash.

Tucson, Ariz.

(CX 5706, Second Revision; IDF 118) (46) Respondent argues that this summary of average discounts does not accurately reflect the real world of crash parts discounts. urges us instead to look at the summary it has compiled of a few .0 In thi opinion we Collowstadard accounting terminology whereby "COt" means the amount a buyer pays 'or a product, and "price" mear the amount for which B sener sella a product, Although GM deaJcru commonly refer to their list-les0% COt of cl"Wlh part as "dealer price, we wil avoid using that deeription, in the (perhaps Rin) hope of avoiding confusion.

.. Twenty-four witness actually testified; the testimony of two additionalwitne8 WEI stipulate as being Lmulative. (T. 1546) ., The invoice appear in the record aa ex 2 through ex 5373. The record al col1tans two aeta of document! rich organiz the dam from ex 2 through ex 5373 na follows:(1) summa.ries of each IB' del!Hngs with each GM der, showing the apoifc crash part purchaa, ita list price e.d dealer wholeule price, and the diBunt the IBS eived, exprea as a percentage off list price (fwvi ex 5374 through Revi ex 5699); (2) summa-riet dtota rly craah part purchtl by each 188, showing the volume of craah pa.rt purcha! in term QfGM'slist price at the diier s wholeaJe price, and calculating an average discunt, expreo as !l percentae off list price. 5700X 5701, Send Reviion; ex 570z.x 5705, GENERAL MOTORS CORP.

Opinion 99 F.

instances where some of the IBS witnesses were able to get slightly higher discounts on individual crash parts or certain lines of crash parts. (RAB 9-10) GM lists thirty-two instances where an IBS testified to receiving a discount equal to or in excess of the discounts shown on CX 5706, Second Revision. What respondent does not reveal, however. is that these instances of "excess" discounts were all taken from the very invoices which are the basis for the averages shown in CX 5706, Second Revision. The "excesses" are part of the average, just as are the' many instances in these invoices of a sale for far less than 27% off list. Thus respondent has in no way impugned the methodology of complaint counsel's calculations on the average crash part discounts received by IBS. Respondent next argues that the testimony and documentary evidence reflected in CX 5706, Second Revision, even if internally consistent, should not be extrapolated to represent public injury, but should be used to show only the private woes of the twenty-six IBS witnesses. (RAAB 11) We do not agree that complaint counsel's sampling technique inherently gives an unreliable picture of the IBS situation nationwide. (47) First, as a matter of proof, it would be an intolerable burden to require complaint counsel to survey the universe of IRS. They are entitled to present examples of the situation obtaining with regard to crash part discounts rather than having to paint the whole picture. We note that this procedure was followed in Associated Press v. U. 326 U.S. 1 (1944), in which twenty-six cities were ilustrative of a national situation. See also, The Coca Cola Co., 91 F. C. 517, 617- (1978). Respondent can then challenge the validity of these examples by showing them to be either inherently unreliable or else not representative. As noted above, respondent has failed to show internal inconsistencies in the data for the seven trade areas; as for the matter of non-representativeness, GM states this to be the fact but has not backed up this statement with any comparable evidence of its own. 59 .. For example, repondents got IB witness Britvic t. "admit" that he received an average discunt of32. 1 % on Chevrolet cra.h part during 1975. For an on-the-land menta cacl1!ation of thirten invoice thi is not to inaccurate. (CX 1489, 1491-93, 1495-98, 1503, 1505-7) In actuality, his 8verr"ge disount on Chevrolet crath part during that time woo 30.2%, but that is atil undeniably higher than the overall Cleveland average disunt of 28% for 1975, /I shown in CX 5706, Send Reviion. However, complaint cOUfiooJ did fiat igore the Britvic Chevrolet cr/Ih part average when computing the Cleveland alkrBlIh part disount. (eX 5706, Second Reviion) Britvic s diaunt on Chevrolet crash part is scrupulow;ly factored into hib total crafh pari average diliount of 28. , which il 1110 made up of hi other average diunt.9 of 25% on Buick, Oldimobile and CtUiJBC Jine crash were developed part and 36% on Pontiac line crash part. Averag.:"I all..rash part diuntB for thrlJ other lB in the sae way as Britvic's and all factred into the final Cleveland figure. .. Without listing al! the ways in which GM could ahow that complaint counsel'a choice of trade areas was' gerryandered one, we note that GM produce no IBS witnea from tmde areas of ita own chooing testify the they cOllistently paid 40% off list on an average Qf al1 linea of cnwh pam, Nor did GM produce invoices fro! which such a conclWlion could be drp.WT. lrntead, GM' s rebuttal oonaiste ofsctten!hot testimony, Blmoot entire (Qmtinw 464 Opinion Moreover, there is some support in the record for a nationwide extrapolation of the IBS situation from complaint counsel's seven trade areas. Most important, of cours, is the already-noted testimony of seven GM dealers and part managers as to their standard level of discount. These witnesses represent five locations beyond complaint counsel's chosen trade areas!m To a lesser, but not insignificant degree, support may be found in the testimony of representatives of three national automobile insurance companies (4S)and of three large regional wholesalers, who opined that the average discount received by IBS and GM crash part ranged from 25-30%. (CPF 109)"' While not direct proof of the terms of any dealer-IBS sales, this third party testimony is independently corroborative of the statitstics previously compiled. Its reliabilty stems from the fact that both insurance companies and independent wholesalers made these observations for business purposes. In particular, the insurance companies in part adjust their business relations to IBS according to their perceptions of the IBS' materials costs. Insurance company servce representatives constantly visit both IBS and dealer boyshops. In the context of what discount the insurance company can get, the topic of the cost of crash part often arises: (49) They have gripe, sir. They like to get things off their chests, the boy shops especially . . " Some things you have to tae with a grain of sat but we check into some things, Sometimes it goes as far as they want to demonstrate what they are doing and actually show you an invoice or something. (Rhoads Tr. 1203, 12100. See also Holschen Tr. I669, 1672, 168: Durbin Tr. 1355-8, 1407-0. The IWs' observations were made with an eye towards the profits available in crash part distribution could they enter, and general from GM dealers or pa maners, referr to ocional pricing vartions on one line of CI"Wlhpart, or evnone typ of crash pa. (RF 141; CRF 141) Th testimony il not of lie nature to and cannot outweigh the cafully conatrnct CB prente by complat count. GD The additional locations reprente by GM dealers and managers ar: Haburg, PellyJvania pa(Sutlif; Memphi, Tennl!, (Schaffer, Trbo and Board); Akn, Ohio, (Mehaffey); Indipoli!, Indi(Byd); and Phoenix, April (Dnton). .. The reprentative of the nationa automobile inurace companies and reonal wholesenl and their teimony appear in the rerd 88: (1) R. Holshen, ClConaultat, State Far Mutual lrulUranre Company. 1a on hi eJlperience in hi asiged ar of nUnois, Wisnsin, Mineata, Colorado, Uta, Wyoming, Alaska, Ho1shen tetifed that Idao, and Monta relevantVlIhin, Orn, diuntE on the range J: "Om 12% to about 25%. (T. 1667, 1689) (2) 30%, averag W. Rhoa, Home Ofce Prperty partClaito Direr Hstate Insurance Company. Rhoads tetifed to an overae dilmt of 25%-3%. (T. 120) (3) R. Durbin owner of Materi Dae, The Relionl:e Inure.ce Company. Durbin tetifed to an average of30%. (T. 1347 50, 1407-140) (4) H. Franck, Ower, Abrlive Supplies, Long Bech, Caorn. Franck is on IW currntly "V the stte of Nevad and Coori who hM previously sold in Washigtn and Oron. The bet dmcOImt ha ob8rved is 25%. (T. 639697, 64) (6) D. Fort, Ower, Fort's Export, Chdeaton, South Carolina. Fort ;ifed to an average dilmt of 25% in Southeat Florida, an ar where he conduct businea. (T. 902-) (6) ordan, Prident, Keystone Automotive PJoti, Pomona, Caifornia. Jord tetifed to an overag disunt of , on GM bumpem notionwide, but with thre metropolitan ar (Nework, &Jtimore, PhiiodeJphi) currntly ng oomewhat grter diunbl. (T. 7797-98) GENERAL MOTORS CORP.

Opinion 99 F.

performance comparisons between the IW and the dealer-wholesaler. (Franck Tr. 6400-1; Jordan Tr. 7793-97) 2. What Dealer-Installers Pay In contrast to the average IBS cost of list less 27% or 28%, GM dealers can purchase any crash part for a cost of list less 40%. (lDF 399) Part purchased by stock order (PAD order) receive an additional 5% discount (IDF 119, 190(1), 248), bringing their cost down to 43% off list. A significant number of crash part are ordered on the PAD" and so are subject to the greater discount." Thus, on average there is a 17.7% cost differential between IBS and GM dealer-installers, to the IBS' disadvantae. (Nelson Tr. 14536, 14561- 62; IDF 398) This is more clearly seen with a part having a hypothetical $100 list price: the IBS' cost is $72 to $73 (28-27% off list) while the dealer-installer s cost is $57 to $60 (43-0% off list), depending on method of ordering. (50) Respondent argues that invoice cost is not a proper basis for comparison. GM would have us take into account the costs which dealers incur in performing a wholesaling function for their own body shops. Since there is no record evidence of the real economic costs of transferring a crash part from a dealership s "front offce" to its bodyshop, GM would have us do as the majority of its dealers do and use its suggested transfer price of 25% of the dealer s cost." Doing so raises a dealer s total part accounting cost to 25% off list which shows to no advantage against the average IBS cost of 27-28% off list. Put in terms of our crash part with the hypothetical $100 list price we see:

List $100 Dealer cost $ 60 (40% off list) T ransler cost $ 15 (25% off dealer cost) Total accounting cost $ 75 (25% off list) All this amounts to, however, is a piece of accounting legerdemain designed to lodge a portion of the profits from a crash parts sale in a dealership s parts and accessories profit center" rather than in its bodyshop profit center. The assigned markup figure could just as .. GM part speista hold that a well-run deale1'h.p will order 80% of its cf/.h part "on the PAD", adi 192) However, in actuw practice deaerompl. especially thos near GMPDwarholU, tend to rely more on ad omera, which quaify only fot the baic disunt of 40% off list. (IF 192, 195). The AL endors complaiv counsel's expert'sl:timate that appro Btely 60% ofcl:aah part orders ar "on the PAD". (IF 122) .. The manager of GM's Dealer Busines Mament Depament teifed that 60% to 65% of all G' dealers ns GM's romended tranfer price of 25% of dealer price. Some us an equally arificial tranfer pri of30% of dealer coo.(Vasuez Tr. 11424) U The major functions of P&A Depaments ar wholeaing, lies of w.srnmty part, internal tranfel1 servce depament8, internal tranfel1 to boy shopl and over-the-nter sales. (R 40) . . .

464 Opinion ll be 10% or 40% of the dealer cost, since it is admittedly not an accurate picture of a dealer s transaction cost. (VasquezTr. 11421- 11422; Nelson Tr. 14016; Benston TI. 16173-78) Indeed, before GM started recommending the use of a 25% transfer cost in 1976, dealers (51)used a transfer cost of the actual retail sales prices charged by their body shop, took all their profits at the "front office."6. The artificiality of GM's suggested transfer cost is also apparent from the disparate patterns of new crash part and salvage crash part purchases by IBS as opposed to dealer-installers. Despite the higher total cost of new crash part to the dealer-installer (once the transfer cost is counted), dealer-installers do not turn to the imperfect but cheaper substitute of salvage crash parts, as do their IBS competitors. Since GM does not require that dealer-installers 66 it is obvious that the transfer cost isuse only new crash part merely an accounting fiction: it plays no role in a dealership decision-making. (52) As a final point we note that some portion of the dealership s 25% transfer cost may represent real costs associated with administration of crash part ordering. Even so, for a valid comparison we would need to know an IBS' administrative costs. As these are not available, the invoices remain that most even-handed method of comparing what IBS and dealer-installers pay for crash parts. The conclusion is inescapable: IBS as a class pay much more. Services GM' s selective distribution system for crash part also puts the IBS under some non-price disadvantaes in comparison to dealerinstallers. The major one is speed of delivery. Dealer-installers have five methods of ordering crash part from their assigned G MPD .. Q. Part and accries buys the part for $6. Bendt it over to the mechaical shop. Mechanical shop inWlls it and chargesthe customer $100. Now, what would be theaetup on the books of the part and a.ccriCf depaent for the sae pri IUd purch!l price oftbat part? A. When did thi happen? Q. Pror to Januar I, 1976.

A. The part department would have a sae of $100 with a COt of sixty. Q. Fine.

A. Nothing would happen in the boy shop.

Q. Then afr January, 1976, how would it show upon the books of the two departments? A. You would have.l BIe in the part and accries department. If they were recrding the salea in accordance with our recmmendations of $75 with a cot of $6. And then you wol.dd have a aae of the part in the boy shop for $100 with a cot of $75, So yol. have $Hi gT profit in one department and $25 in the other, a tota of $4. (Vasuez Tr. 1149811499) U The Deer Saes and Servce Agment (R 2) conta no requirement that GM detdere intal only new Ih part; nor do we fmd any teimony in the rerd of auch a requirement. " Accrdg to roznplaint count' s expert, the only adinistrative C0 8!iate with a dealer s orderig II for hi own boy shop ar the minor cot of looking up the part number and ruling out the order form. This , at most, 2% ofthe dealer s saes price. (Nelan 'I. 14017- 18, 14020, 14559) One part maager for a ieaership state that filing, out crash part ordel" is probably the eaiest administrative tak of a dealer depaent. Qknton Tr. 120) GENERAL MOTORS CORP.

Opinion 99 F.

warehouse, four of which provide shipment of the part immediately or within 36 hours after the order is placed. (IDF 190, 191, 193) By contrast, IBS, even if located next door to a GMPD warehouse, must obtain the part from the appropriate franchise dealer. If the part is in stock at the dealer, an IBS may be able to have it in a few hours; but the normal delivery time, even when the part is in the dealer inventory, is 24 hours. (See, e. Daniels Tr. 2451-52; Niebling Tr. 2517-18; Clouatre Tr. 3773-75; Trepagnier Tr. 3841--2) The wait is much longer, however, when the dealer must order the part, which happens frequently. When this happens, even in the GMPD warehouse cities of Buffalo, St. Louis, Cleveland and New Orleans, IBS commonly experienced delays of three days to a week in receiving parts ordered on their behalf by a dealer-wholesaler from the local warehouse. (Daniels Tr. 2277-78; Clark Tr. 4151-52; Britvic Tr. 7503; Trepagnier Tr. 3841--3) Moreover, the IBS does not receive the part directly from the GM warehouse, as does the GM dealer; the part going to the IBS must be routed through the dealer-wholesaler. (53) IBS may also be disadvantaged on ordering and supply convenience. A GM dealer generally goes to one source for all his crash part needs. If his assigned GMPD does not have the needed part in stock it undertakes to find and arrange for delivery of the part through computer link-ups with all other GM warehouses and, if need be, GM itself. (IDF 176, 180) By contrast, since dealers tend to wholesale only the crash parts applicable to their franchise," an IBS has to contact many different suppliers: the Buick dealer for Buick crash parts, the Cadillac dealer for Cadilac crash part, etc. Not having a "one stop supplier" means increased delay and administrative expenses in ordering crash part, as well as decreased abilty to bargain for the best price or establish credit terms.

Q. I would like you to compare the advantages and/or disadvantaes of being able to purchase products for all makes of car as opposed to having to go to various sources of supply for each make of vehicle.

A. I would think it would be an advantage, deliveryise. I would have less calling. I make anywhere from 20 to 50 calls a day callng various dealers in supplying myself where I could limit that down to maybe 5 calls instead of 20 to 50. I think it would be a lever in getting better servce if one man had all my business in your GM line. (Latakos Tr. 4034-35) (See also Perschall Tr. 3164-5; Franck Tr. 6413; Niebling Tr 2594) However, when it comes to delivery terms, IBS and dealer installers seem to be on roughly equal ground. GM routinely prepay U The Wholesale Compensation Plan is not, as might be BUlpete, reapIn.ible for this tendency. Even bef! 196 dealera generally wholesed only their line of crash part. ffid 34; SutliffTr. 11025-26; Perkit1B Tr. 9871voo FEDERA TRE COMMISSION DECISIONS Opinion freight expenses on the majority of its dealers' orders. (IDF 190) There is no such nationwide uniformity of easy credit terms on delivery for IBS; on the other hand, it appears that many dealerwholesalers provide free delivery, at least locally. (Rhoads Tr. 1200; Bogard Tr. 10477-78; Mack Tr. 11570-71; Denton Tr. 12003-4; Tribo Tr. 10827; Baker Tr. 6193-94; Daniels Tr. 2277; Barney Tr. 5240; Serwacki Tr. 4902) Some even absorb freight charges on widearea deliveries, if the order is average-sizd or larger (IDF 117) (54) Return policies show a similar lack of major disadvantage to the IBS. A dealership s part return privileges allow returns for any reason, up to certin financial limits. 6. The IBS have no assurance and the record does show that some dealer-wholesalers charge a fee for returns. (Wicker Tr. 5560; Hershey Tr. 6563; Weatherford Tr. 6789-90) On the other hand many other dealers provide liberal return privileges, some even without charge. (Smith Tr. 7419; Britvic Tr. 7506; Neal Tr. 7937-38; Finkle Tr. 9434) In sum, the record does not show that IBS consistently receive less accommodation on delivery terms than do their competitors, the dealer-installers. Finally, we conclude that GM's selective distribution system for crash part does not withhold from IBS any vital technological assistance. In the first place, no great amount of technological expertise is necessary for performing body repair work, as compared for instance, to the repair of a vehicle s emissions control system. Secondly, GM dealers themselves receive almost no technical assistance from GM concerning crash part installation. (Murray Tr. 10019, 10021, 10030) (c) Effects 1. Insurance Work The record does not show precisely how IBS businesses are affected by any nonprice disadvantaes on crash part delivery. There is of ,ours testimony confirming the obvious fact that easy availability md comparable delivery costs are important to an IBS' ability to ompete in crash part installation. (See e. Perschall Tr. 3273; erwacki Tr. 4875; Clark Tr. 4176; Smith Tr. 7474-75) The price Bcrepancy, however, is clearly key to the problems IBS face in 'tting insurance-paid business. (55) Approximately 90% to 95% of all crash repair business done by dyshops, independent and dealer, is paid for by insurance compa- Return . Deei' ca return pa for crt and without chae up to an amount caled the Maum rye. The Maum Return Rerve is eared by the dealer at the tate of eight percent of stock ("PAD") order initia order and "quaifed speia merchadising orders ) and four percent of supplementa utok ordera, rs ar chaed B penalty of 20% for cetuma in exOO of the Maximum Return Rerve. (RPF 33) .. g., GENERA MOTORS CORP. 595 Opinion 99 F.

nies. (IDF 200) In the interest of cost control, most major insurance companies have adopted a system of recompense which allows the insured to go to the repair shop of his or her choice, but sets upper limits on the amount of the claim. If the cost of repair exceeds the insurance company s limits, the claimant must pay the excess. (IDF 201; Holschen Tr. 1690; Hershey Tr. 666!H7; Rouse Tr. 7026; Britvic Tr. 7547; AlbertinTr. 8286-7) Obviously then, car owners have great incentive to go to boy shops which can meet the insurance company s appraisal for repair of crash-damaged vehicles. This incentive is reinforced by the fact that most insurance companies provide theirc1aimants with lists of "preferred" one-stop competitive" bodyshops, meaning shops which have ageed in advance to accept the insurance company s estimate of repair cost. (IDF 205) A very large portion (40%-80%) of insurance-paid work is performed by bodyshops to which the claimant is referred in this manner by the insurance company. (Weatherford Tr. 686 Rhoads Tr. 1217; Hoschen Tr. 1690) Getting on the "preferred" list, or at least being able to do repair work within the limits of an insurance company s estimate, can therefore be crucially important to a bodyshop s ability to stay in business. There are three elements of a repair appraisal which can be adjusted by a boyshop to bring the overall final cost of the job within the estimate: time, labor rate and crash part discount. The last is by far the most significant.

Insurance company appraisers use standard Hcrash manuals" to determine the time needed for each typ of repair job. (Durbin Tr. 1449) Payment is made on estimated rather than actual work time. Thus, if a bodyshop can consistently "beat the book" e. do the job in less time, it can use all or a portion of the amount (56Jrecompensed for unnecessary hours to provide leeway for adjustments on the other two elements. (Witman Tr. 5017) However, both the natural bounds of labor effciences and the shop s desire to do quality work limit how much time can be shaved on any job. (Witman Tr. 5016) Witnesses uniformly spoke in terms of lopping "an hour or two" per job. (See, e. Serwacki Tr. 4836; Whitman Tr. 5016-18; HersheyTr. 6629-30) This does not translate into particularly high savings. For example, at a $15.00 per hour labor rate 71 a heroic 25% reduction 19 Th iI a gentleman s agment" rather tha a formal contract; adustmentl in the rmal figure can be negotiate. However, Buch negotitioIU generally tae plac afr repair' have ben and are limited to items overlooked by the atUBrs. It isnot 8Urpriin, therefore, that one inurace oomPMY :rprentative estimate tht 75% to 80% of all este ar acpt without chae by "preferr" mopi. (Roa Tr. 1221- , 1322) Ia.borral: for 1978-1979 Th W8 e. 1978 rate reference by Buffalo witness. In the other trade arraed from $11.0000$16.00. TuCtn: $11 (Brokaw Tr 6029), New Orlea: $12 (Tpaier Tr. 3978-79J,Spokae: $14 (Rus 'f. 702) field, Ohio: $14 (Hershey Tr. 6611-12), St. Louis: $15 (Diela Tr. 23). 596 FEDERAL TRADE COMMISSION. DECISIONS 464 Opinion 12 produces only a $30.00 saving.in time on an eight hour job This brings us to the second means of m eting an estimatediscounting the stated labor rate. Insurance companies use the prevailing" or Hgoing labor rate for the area in writing estimates. (Tr. 1450-1) Theoretically, a bodyshop could offer a lesser rate to balance higher part costs or longer than average time per job. Unfortunately, in many areas labor rates are simply not adjustable being negotiated by unions and gen rally followed by even non-union shops, IBS and dealer-installers alike. (Durbin Tr. 1440; Clark Tr. 4185-6)12 More importantly, given their relative importance as components of the overall claim, it would take a massive adjustment in the labor rate to offset a minor discount on crash parts. Several IBS witnesses gave variations on this example: to replace a front end takes 8-10 hours of labor; reducing a typical $14.00 per hour labor rate by 10% results in a saving of $11.20 to $14.00 on a front-end job; but since the aggregate cost of the relevant crash parts (headlights grile, bumper, fenders, hood radiator support) ranges from $1200 (Chevrolet) to $2200 (Cadilac), a reduction of 10% on parts produces a savings of $120-$220. (Serwacki Tr. 4836, 4841; Barney Tr. 5364- 65; Whitman Tr. 5013) (57) This shows why the third means of meeting the estimate discounting crash part, is paramount. Insurance companies uniformly demand a sizeable discount on crash part. Such is .their buying power-as the ALJ noted, insurance companies are the real consumers of almost all crash part (IDF 200)-that no body shop can get on a preferred list without acceding to the crash parts discount prevailing in his locality. (IDF 206) In each of complaint counsel's seven trade areas we are presented with a picture of GM dealer-installers consistently giving at least a 10% discount on all crash parts and frequently more, up to as much as 25% on certain parts. (IDF 208, 209) In contrast, the IBS often cannot give any discount, rarely manage more than 10%, and that only with diffculty. The following chart summarizes the testimony on comparative crash part discounts in the seven trade areas: GM Dealer- Area Installer Discount IBS Discount Source Buffalo, N. 10-25% 0-10% IDF 209 Cleveland, Ohio 10-25% no more CPF 138: than 10% IDF 208 .. The tiroerequired for a full front-nd replllcement (Serwacki Tr. 48). Many other jobs reuire Jes extensive repara, and con5luently offer lea IIpe for RBvingl throughspe. TO We doubt that any boy shop owner wolild daideny the prevailin Jaoor rate to his "heavy hit man (Serwllcki Tr. 4829) GENERAL MOTORS CORP.

Opinion 99 F.

GM Dealer.

Area Installer Discount IBS Discount Source Mansfield, Ohio CPF 139: IOF 209 New Orleans, La. 10-20% 0-10% IOF 209 St. Louis, Mo. 10-25% no more CPF 138: than 10% IOF 208 Spokane, Wash. 10-20% 0-10% IOF 209 Tucson, Ariz. 10:25% no more CPF 138: than 10% IOF 208 The overall amount of insurance-paid business lost by reason of the IBS' failure to match dealer- installer crash parts discounts is not quantified, and being in the nature of an "if (58)only" probably cannot be." However, the record is replete with testimony that such business is so lost; the witnesses include insurance company representatives as well as IBS operators. (IDF 211-212) The following are typical:

Our margin of profit is detennined by two things: labor and price discount. The price discount is what gives us our lifebloo, and when they take some of that away from you, they re taking some bloo away. . . . If they gave me 10 percent, I would have stil been in business toay. I stil want to fix cars. I love cars. It' s everything I've ever done all my life. I couldn t fight it anymore. I can t work until 3:30 in the morning anymore. I don t want to be found dead at 2:00 in the morning alongside one of my cars. (Serwacki Tr. 4869-70) I am talking about the one and two-man shops that work many hours a day, a lot of evenings and have the highschool kids or whatever coming in and helping them evenings and on Saturdays. A lot of those disappear and go away. (Daniels Tr. 2310) The examples in the record are suffcient to support the proposition that the IBS are substantially harmed by the cost discrepancies which stem from GM's selective distribution system for crash parts. (59) 2. Statistical Proof To go beyond the examples cited above of the IBS' lost insurance work in search of statistical proof of harm to the IBS is to labor in 8 TO We know that 36% of 100 reenues come from sae of all makea of crash part. (CPF 122) We also kno that in the crah part univers the ratio of GM crash part to other typ of crash pam corrponds to the ratio GM cats on the road aa compare with other makes: in 1976 that ratio was 45.5%. (CPF 123) Thus, if IB could r all the GM craah part they nee, we would expet GM crah pam to account for 16.38%-24.57% of IBS revenr from part saes. A lower percntae would indicate inabilty to get GM crash part. However, we cannot il Ruch calculations beu! the reord nowher break out the percentae of IBS revenues attributable to sae of GM cr8!h part.

However, we should al note that a Rhowing thatlBS achieved 16.38%-24.57% revenuea from saea of neVI crah part would not necly prove lack of dirimination. The starting fJ.re for theo calculations (36 revenues from Haes of all makes of craah pam) has a built-in bia, reflecting a world where no maj( manufacturer wil sell its crash part directly to IB. 464 Opinion barren field. Whittier summed up the problem: "For all sad words of tongue or pen, the saddest are these: 'it might have been!'" More prosaically: it is impossible to gauge the amount of business lost due to a refusal to deal which has not been preceded by a course of dealing. There is no yardstick against which to measure the loss, no before and after " no minuend from which to subtract. If this were a private cause of action for damages resulting from a refusal to deal, the inability to quantify harm could be fatal. However, even in private cases where damages are too (60jspeculative to support recovery, court have acknowledged that proof of the fact of injury is an entirely different matter from proof of the quantum of damage. See, e. , Zenith Radio Corp. v. Hazeltine Research, Inc. 395 U.S. 100 (1969); Fleer Corp. v. Topps Chewing Gum, Inc., 501 F. Supp. 485 (E.D. Pa. 1980). Section 4 of the Clayton Act, which requires proof of private injury as a prerequisite to a damage recovery, is of course not imposed upon the government which generally proceeds to enforce the antitrust laws by suits equity, undertaken in the public interest. Thus in all the cases previously discussed which produced a duty to deal order, nowhere did the government have to measure the injury to competition with statistical precision. For example, in Otter Tail Power Company Co. v. United States 373 U.S. 341 (1963), the refusal to deal, on its face supported equitable relief. The Court inferred that municipalities ,. The term "yartick" and "before and afr" ar borrwed from the jargon of computin loot profits in private acioll. The yartick theory relies on an imalogOWl cles of unhared competitore whose saea daeduring the daag period ca be us to approxiate what the plaitif could have achive. The before and afr thoory looke to the plaitifs own sees hitory just before the supplier bean ro refus to den and extrapolate from that. Neither theory ca be us here. The IBS have nlWf ben able to buy cr8.h peru direly from GM; conauently, they ca furnieh no prior sae! hitory. The dealer boyshops ar the IB' only competitors, and whi they ar similarly aituate enough to allow the crllh pa invoice cot compmnsdesribe earlier, their position IU a pa ora nmeb larer buaine8 preludes meagful comparins to the IB on profits, growth and failure.

to meaaure the TO In the private acion arna erlfUive debate exists over what deg of prof 18 necleg of ha caus by an antitn.t violation. Se e.g. Han, Th 1mt Prfil8 Mt'ure of Dnma in ':IIMncment, 64 Min. L. Rev. 751 (198), L. Sullivan, Hadbok of the Law of Antitrut Seion 251 at 785 977); Giban Th "Market Skare" Thry of Da1T in Prvate Enforcement Ca, 18 AntitI't Bull. 743 (1973), lloil Dama Detfrminatio in Pruate Action Suita 42 Notre Dae Law. 647 (1967). The underlying issue is, of 1!' , how to achive a balance betwn a atron. judicial policy again apeulative dllag, and a diinclintion favor the wrngoer whose wrng doing is the moet effecive and completethe latter premis upon the ,wled that "the marketplace usually denies U3 sure knowlede of what the plaitiffs position would have bon ,nt the defendant's antitrut violation.

. Ofoou1", even in the private arna, when dwag ar either unavaiable or not the mOBt effecive remedy, :able relief is somethnff :available.Se, Antitrut Violations Without Daage Reveries, 89 Ha. L. 1127 1139 (1976).Se also &rkey Photo 1m: v. Eatman Kodk Co. 60 F.2d 263, 292-93 (2d Cir. 1979),cert. 1, 44 U.S. 1093 (198), where plaitif photormi9her oought only equitale relief for didvantaes stemming 'lod' s policy vi sellin only liite ru of color paper (a vita photoni9hin s\lpply) to its competitor. inisher. Kod ha B 60+ % shu vi the color paper prouction market, but only 10% of the photofmi9hig . The court quickly concluded: "given Kod' s monopoly power in color paper, tbi refus to dea would, uatifed by a valid busines n, appe to violate!2 and form the basis for equitable relief.1d. at 292. I waa remanded for furher fmdi on thi issue but settled without rmdi. GENERA MOTORS CORP.

Opinion 99 F.

that could not obtain vital supplies for independent power systems were injured by the ensuing lack of competition in power retailng. Otter Tail relied upon United States v. Terminal R.R. Ass ' , 224 S. 383 (1912), which went even further: fear of potential injury to competition motivated the court's equitable decree. There was no showing that the Association had yet used its monopoly of approaches to St. Louis to impose discriminatory rates upon nonmembers; the court merely noted that the possibilty was "inherent in the situation, uplain, uundeniable" and Hohvious . 224 U.S. at 397 400 401 405. (61) In Associated Press v. U.S. 326 U.S. 1 (1944), the By-Laws of the major U.S. news agency prohil-ited members from sellng news to non-members and granted each member powers to block its nonmember competitors from membership. The Court, without requiring any data on lost business held that the By-Laws obviously hindered and restrained the growth of competing news agencies and non-member newspapers:

It is apparent that the exclusive right to publish news in a given field, furnished by AP and aU of its members, gives many newspapers a competitive advantage over their rivals. Conversly, a newspaper without AP servce is more than likely to be at a competitive disadvantae. 326 U.s. at 17-18. . In Grand Calliou Packing Co. 65 F. C. 799 (1964), the fact that West Coast canners were injured by respondent's discriminatory leasing system was determined by the Commission on the basis of (1) calculating the cost differential between Gulf and West Coast canners and (2) accepting testimony of individual West Coast wholesalers and processors that they were losing money trying to match the Gulf Coast prices, but could have matched these prices had their lease terms been equal. 65 F. C. at 835-36, 841-845. The opinion did not utilize statistical comparisons of the overall growth of West and Gulf Coast canners as a clas. In affrming La Peyre the Fifth Circuit was even more spartan in its determination of competitive injury, relying wholly upon the rental rate differential Lapeyre v. F.T. 366 F.2d 117, 120 (5th Cir. 1966). (62) In our recent decision in Reuben Donnelley, continuing injury t the commuter and connecting carriers as a clas was premised up' an inference of lost business,'. rather than proof of the ex, TO Indee the Court gave only ana "ilWltration" oChow the By-Lowl! worked to block new entry and grw non-member competitive neWlpapers, notin that in 26 cities of the U. S. exitin newspapers alredy had con! for AP newly and ha contract with rival news aernce which severely limited any new paper lI acc to the 326 U.S. at 13.

(Sjeventy percent of commuter caenl pBngni are connecin to or from certifcate caen the failure to list connecting flight information for commuter carrenl deprived them of 8 r marketin to! with repre toe. large portion of their blUel. 95 F. C. at 55. (Co , 464 Opinion amount lost, down to the last decimal point. In re The Reuben H. Donnelley Corp. 95 F. C. 1 (1980), rev d on other grounds 630 F. 920 (2d Cir. 1980).

Finally, even in the private action of Gamco, Inc. v. Providence Fruit Produce Bldg., Inc., 194 F.2d 484 (lst Cir. 1952), the fact of plaintiff's exclusion from business opportunities associated with the best-situated market in town, without more, was enough to impell the Court to order that equitable relief be awarded on remand. 194 2d at 489. See also Zenith Radio Corp., supra, 395 U.S. at 132-133. The conclusion is inescapable. Certain antitrust violations, usually those particularly susceptible of correction by government action create a situation where the most effective relief is at equity, not law. When such cases involve an ongoing refusal to deal, harm to competition can be proven by testimony about the experience of representative members of the discriminated-against class of (63) competitors. Statistical surveys of the class' competitive strength, if closely related to the marketplace affected by respondent's actions would be a welcome supplement. However, if the statistical proofs cannot be finely enough drawn, their absence is not fatal to the case. With this in mind, we review the statistical evidence available on the record.

A vast portion of this record is taken up with argument concerning the proper measure of the IBS' competitive strength. Their numbers exits, sales, and profits are measured by means of at least five different data bases over various time spans.' and compared with the same attributes of dealer body shops all servce industries " or selected retail business, as the advocate s convenience dictates. In these comparisons, respondent attempts to show that the IBS are )etter off than assortd other businesses; complaint counsel of course '.ttempts to prove the obverse. Generally speaking, respondent lVors the numbers and sales data which show a growth trend in the 'erall class of IBS. Respondent argues that the increased numbers (Mt WIrs of the OAG read the Jiatinga of the flighta between n city-par from top to bottom and pick the first convenient flight; therefore, listing the flight/ of certificste carners before the flghta of noncertificste carriers oftn re9ults in WIrs pickin a certificate flight without even looking at the liatingu for noncertificate caers. 95 F. C. at 83. Indee, the Send Circuit specificaly endors the Commision s methodolog of determining injury and 1clusioll ba theren. 63 F.2d at 92425.

One qu.. which touchea each act of awtitica concems the cprrect time period to examine. Generally 'g (for neither side scruplesto be mcoruiBtent for the Bae of a momentary advantage), complaint counsel 1 the years 1972-1976 whereas pondent start ita trending in the mid-siJties. Nearly every mellure Ie In enterig upon leaer times in the early 1970's; respondent naturally wants to rai the average by 1 the more properoWi earlier yea, whie complaint CQunal wanta to cut thos sae years out, both picture WI gr as poible and to avoid any implication that the IBS reached the naturallimita of their .fore 1972. To WI, both starting points sem equally capricious and dictate more by thechoi of data !iY defined relationship to crash part a.vaHability, GENERAL MOTORS CUIH'.

Opinion 99 F.

are a guarantee of increased competition. Complaint counsel rely more on exits and profitability data, which show that many individual IBS are unsteady competitors. Complaint counsel argue that competition must be assessed in terms of quality, not merely quantity.

(a) Growth in Numbers There is no question but that, overall, the number of IBS increased between 1963 and 1977. Various subgroups of that class have been measured over various periods of time in that (64Jfourteen-year span by four methods of varying accuracy."' The bottom line, however, is always growth-by at (65Jleast 40% over the 1967-1977 decade. At the same time, GM records show a slight decline (4.2%) in the estimated total number of dealer body shops. (RX 38)"8 A simple comparison of growth rates, however, does not translate into an accurate comparison of competitive strength. There are too many different factors affecting the ease of growth for IBS and dealer body shops. The "independent" in the term Independent Body Shop is truly descriptive: and IBS can be started with a minimum of equipment and space and is often a one or two person enterprise. "' (1) Ce1UWi Data: The number of IB. (SIC Coe 7531) reporting to the U.S- CeMuS Bureau increas by 28.5% (4,621) between 196 and 1967 and by 52.7% (10,982) between 1967 and 1972. (IDF 26, 26) No CenaUl data are available for years after 1972, which curtila the usfulnes of thi Bet or statistics. (2) IRS Data: Th data is limited to IB partnerships and proprietorahip$. It shows a numerica increao, in percentae term, of 51% from 1967-1972. Between 1972 and 1976 the growth levels off abruptly, 6showing an incrementa increll of only 2.9%; but this stil provides II g.owth of 55.3% for the years 1967-1977, during which dealer boy shape declined in number by 4.2%. (IF 26) Giru:e IBS corprationB account for approximately 10% of the tota number of ID busines (Nelsn Tr. 13789, 13796), their excluaion probably doe not skew thi data utlduly (it wil be a diferent story on BBes and profit trends derived from IR data), (3) Telepho DilYtory Li.ting: One of respondent's employee prepare a tabulation (R 41) purprting to show the number of ID in each of compla.t counol' s selected trade areas (except Manfield, Ohio) for the years 1967, 1972 and 1977. The numbem were arrve at by the simple prDC of counting the listing utlder the Automobile Boy Repairing and Painting" heading in the clllifed ad seion of the phone books for the greater metropolitan aras of each city, in each year. Although the ALJ flm6 to find that such surveys ca be accurate adf 278) we conclude that thia survey, at leWlt, is riddled with methodologca errore. For example, respondent made no attempt to veriy that each liate company WWl actu.ly ill bUlines, was engaged priarily in boy work (WI oppo to pat work, mechanical repare or salvage work), or generally worked on GM vehicles (as oppo to foreign auto). (Stoker Th. 11334, 1134, 11351-5, 11369, 11376-77). Even more serious ia respondent failure to Be if the companies liate in the earlier phone books were stil listed ill the later years of the tabulation. By complaint coun.l's calculations, fully 28%-4% of the compaies listed in 1972 did not appear in the 1977 Yellow Pages. (CR 167) (4) Dun and Brotret Dato.The number of lB surveyed by D&B grw from 11,64 in 1972 to 1786 in 1977 (+ 53%). Since D&B surveys only thoo bo shops for which it ha ben requeste to furnish a cret report (Brr Tr. 12141 145) thia aaple doe not tell us anything about the overall growth in numbers of IBS, but merely pictl.resthe ri in numbers orms whoacret ratingwaa of interest to a D&B client. U The GM franchis agment reuire dealers to submit a Trl Ballcc Form (Financial Statement containg 8. variety of accounting da.ta to the Dealer Businea Management Department; about 90% of the dealer dQ 80. (VasuezTr. 1l410-13, 11432) The Tr Balance Form shows aaes from the boy shop and thus how many ( the reporting dealer1hips have boy shape. Of the 10% who did not report, GM estimate that the percentae' nonreporting dealers with boy shnps exactly corresponded to the percenta of reporting dealers with boy sho( (VB8uez Tr. 11433) Complaint counol claims thi "blowup" method of determining the univers of DBS unreliable, but we think 8 10% or 1- errr ia accptable in thi context. We note that we sanctioned "blowu methodolog in Retail Crit Co., 92 F. 131, 140(1978). oul: FEDERA TRADE COMMISSION DECISIONS 464 Opinion (IDF 264) By contrast, a dealer body shop is always an adjunct of a GM car franchise s, and establishing a car dealership involves capital outlay, personnel commitment and other requirements - not the least of which is fillng a need in GM's automobile marketing plan - unknown to the independents. The numbers indicate only, therefore, that entry into the independent body shops business is perceived as desirable and is relatively easy; they tell nothing of the level of competition offered once the low entry barriers are breached or the abilty of individual IBS to stay in the business they so easily entered. For these purposes we must look to the statistics on exits. (66) (b) Exits Respondent maintains that IBS have a low failure rate in contrast to many other retail lines of business, relying on Dun and Bradstreet report which state that the IBS failure rate has declined from 32 per 10 00 in 1972 to 15 per 10 000 in 1977. (RX 303B) This represents the next-toIowest failure rate for any of the twenty-three retail lines of business for which D&B keeps failure rates. (RX 303A) However, the D&B report are so flawed as to IBS that cross-industry are impossible. As noted in note 82comparisons, even if valid,85 supra p. 64, the D&B universe of IBS is not representative of the overall IBS universe. Moreover D&B' s definition of "failure" is too limited: It excludes all firms which cease business but leave no unpaid bils IBS which operate on a C. D. basis, or which simply become discouraged at continuing low profits and wind up the business honorably. Finally, D&B's collection of failure information misses all personal bankruptcies, a course of action open to the 90% of IBS which are proprietorships. (Nelson Tr. 13774-75; Wyant Tr. 12236-1) In contrast, complaint counsel's statistics, based upon Census data for 1972 seem fairly reliable. They show a failure rate of 1 140 per 000, the fifth highest of the 24 retail lines of business. The Census Iniverse is more truly comprehensive, as is its definition of exit from lusiness. While the cross-industry failure rate comparison remain s tenuous under Census-based calculations as under D&B statistics, 04 Sigifcatly, the 1967-1977 decline in dealer boy shape i8 slightly outpace by a decline in deaerships . indicatin that the decline in dealer boy ohopa may be the result of overall fr&1china matters, rather tha lpetition in boy work.

.. The fret1coating varbles which complicate compans beween IB and dealer boy shop€ ar even more \emus when compa IB to Department Stores" or "Lumber & Buildi Materi"--d the nexus to 1. pa is even more tenuous.

, The Census Survey of Selec Servce Industries publishee thenumber of busines which were in busing the year but have busine6 by the end of the yea. The ext data (i.e. ar you in busines" io ce !lte by the survey fonn. which must be a.were under oath. (Nelsn 'I. 137657)The form of the queetion d to aslight overstatement of failure rate, WI it picko up ms diBntinuence for other tha total failure (D.mtinurJ GENERAL MOTORS Cu,,.

Opinion 99 F.

and for the same reasons, the Census figures do have limited uses. First, the Census data (67)casts further doubt on the reliabilty of D&B statistics. Second, for one year at least it corroborates the IBS witnesses' testimony about their diffculty surviving. The necessary link between crash part and business failures is provided by that same testimony, which shows that the cost of crash part is a crucial variable in a body shop s profit picture. (c) Sales The record contains a welter of fragented sales statistics for dealer body shops and IBS in three distinct time periods between 1963 and 1977. Owng to the use of three disparate data bases, only for 1967-1972 do we have data covering the complete universe of IBS and dealer body shops..' That data shows that aggregate sales of IBS increased by 116% while aggregate sales of dealer body shops increased by only 40%-48%. .. (68) The fact that IBS aggregate gross sales increased at over twice the pace of dealer body shop sales does suggests that the IBS are strong competitors in repair of crash-damaged GM vehicles. However, it is important to note that "sales" includes, for the IBS especially, far more than auto collsion repair work: as described more fully under the "Profitabilty" subheading, infra over the years, due in part to inabilty to compete in body work, the IBS have diversified into other lines of repair work. A significant, though unquantifable portion of the IBS increased volume of business is generated by these sideline operations. (IDF 282) (d) Prfitabilty The record does not permit any comparisons of IBS and dealer body shop profitabilty,.' since it lacks any information on dealer notably chage in legal form of the orgiztion (i. , from a partnership or proprietorship to a corpration). However, lIince only 10% of IB are corpratione (NelBn Tr. 13796) we ar inclined to think the bias a8light. n For 1961967 we have Oenauo data on the agate sales of both IB partnenlhips and proprietoI'hipa and IB OOrpIltiOIU (i.e. the complete univer& oral) anf 269), but no data on dear boy shop aaes. For 1967we have GM estimate on its dealer boy ahops' agate lie! (lDF 270) and both CeIlUB data on all IBS saes adf 270) and IR data on agate saes of IB proprietnlhips and panerBipe (i.e. IB corprationa excluded). (CX 781S-7819) For 1972-1977 we continue to have GMI tftimate on agate 88les by dealer boy 8hopa (R322) and IR date on agate saes by IBS proprietol"hips and partnenlpa (CX 7818-7819), but the lack continuin CeIlUI data mean there is no dire prof of corporate IB uael. The qUeltion of whether such pro ca be reched by an estimating pro known as 'linking' W.l battled hotly throughout the trial and appeal. (& IDF 272-273) Since we conclude that the gr saes fIgre ar not suffciently fQCU! on crlUh part to be of \I in 111i analysis we do not nee to cofijder the iBue. We do note however, that even under the incomplete ae data mot favorable to complaint IXHmatd, thegrwt in ID sees volume for 1972-76 WIU only minutely les th, growth in dealer boy shop BOes (I panership! and proprietorships: up 37.1%; dealer boy shape: up 37. (CX7819) U Complaint coulll arge that GM'a estimate on dealer boy IIhop 8Bles areundenrte by 20% be five cateori of dealer boy IIhop reenues were systematicaly e1lcluded. (CR 157-161) Even if thi 201 facre bak in, however, the cotrWlt between IB and dealer boy shop aggrate saes growth is atil grat. .. Once again, we reist the invitation-thi time respondent'lIt. us thes statistic! to make Croindl (Cmdi, 464 Opinion body shop profits. IBS partnership and proprietorship earnings as reported to the IRS for 1967-76 are summed up at CX 7818-20. The data is, like the other IRS statistics, incomplete, as it does not include IBS corporations. This definitely skews the low profit averages which complaint counsel draw, since IBS corporations being generally the largest, most profitable IBS businesses, produce about 32% of total IBS business receipts. (Nelson, Tr. 14301) Nevertheless, since IBS corporations represent only about 10% of IBS businesses by number, the IRS statistics are accurate as to approximately 34 561 IBS partnerships and proprietorships (1976 figures - CX 7819; 7820) (69J-a subgroup of no little size. As a class non-corporate IBS did fairly well from 1967-72 with aggregate net income increasing by 54%. (RX 314A) (We have already noted the increase in IBS sales during this period.) However, in the years 1972- 76 non-corporate IBS aggregate net income increased by only 3.3%, while the percentage of non--corporate IBS losing money jumped from 7.2% in 1967 to 19.9% in 1976. (Derived from RX 315B; RX 320) On a per shop basis and in stark dollar terms this means that the average net income of an IBS proprietorship/partnership was $5 256 in 1967-and only $5 410 nine years later. With the consumer price index for all services increasing by 180.4% during those nine years these IBS needed to be making $9 482 in 1976 just to stay even in real income with their 1967 earnings. (CX 7820) The profiability statististics are derived from IRS Schedule C, a form which is fied to obtain a variety of deductions allowed in the tax code. Accordingly, Schedule C fiers have an incentive to overstate expenses and, possibly, understate revenues on the form; such inaccuracies would be reflected in the a.ggregate numbers. (Bentson Tr. 15841--3) Respondent's expert was unable to quantify or even guess at the extent of this bias. (Id. On the other hand, the record contains testimony which supports the inference drawn from these statistics: that many IBS have faced lecreasing profits in more recent years. Witnesses observed that IBS 'owners often cut corners by paying themselves low salaries 'working extraordinarily long hours, and employing their families at We or no remuneration. (IDF 280; CPF 155, 161, 176-179) Even so any of the IBS witnesses from the seven trade areas, testifying as parinshere of IB with Genera! Auto Repar Shops. 21) neral Auto Repair Shops Bpeializ in (R riea and mochaical work, for which no craah part are require and for which parta are readily available a multitude of auppliern. Genera. Auto Repar Shops (GAR) tle analogoU9 to dealers' mechanica shop!, not f boy shops. (Nelson Tr. 1522426, 1430091) , e.. Daniela Tr. 2292, 2293: yearly 6lafy of $17 00 over It number of years for 60 hourn of work a Lato Tr. oWIO-U: yearly saary ofapproJtOna.te\y $15 00 during 1972-1976. Similarly, former IBS owner -d Serwacki obarvoo (T. 481-42, 489) owners of "many newer shop!" working "all kinda of hours" for \150 a week in order to stay in bllinea- (Serwacki Tr. 481--2, 489) GENERAL MOTORS CORP.

Opinion 99 F.

to the years 197&-1977, told of experiencing losses or significant decrease in profits. Moreover, some portion of the small profits (70) which these IBS were able to eke out comes from increasing diversification into non-crash work such as towing, office furniture and cabinet repair, mechanical and radiator repair, frame straightening, customizing and specialized (non-car) collsion repair. (IDF 281-283) IBS are doing less work on GM vehicles, as a percentage of total business, than formally (IDF 247, 282) Even as to the GM work remaining, some is sublet work from dealers. (IDF 284) (e) Conclusion Our conclusion, after a careful examination of the statistics in this record is that the parties have pretty much battled to a stand off. None of the four proposed measures of competition performance allows us to make direct comparisons between IBS and dealer body shops: the difference between a body shop dedicated solely to repair work and one which is part of a new car sellng operation introduces too many variables into comparisons in terms of growth in numbers and profits and exits; also, the uncertainty of what products are included in sales records obscures that comparison. Moreover, both complaint counsel's and respondent' s chosen data bases suffer the same fundamental flaw with regard to the focus of this case. Each yields some inference about the competitive vitality of GM dealerships or the IBS, but in no case is there a measurable correlation between the availabilty of crash parts and the overall competitive condition shown.

The most we can use these data for is to examine the IBS' competitive health, standing alone. Here we think that the statistics on failure rates and profits for non-corporate IBS enlarge on the testimony about individual IBS' diffculties in maintaining their businesses. The data therefore confirm that the seven trade areas are fairly representative of national trends. When these data are read with the testimony and document described earlier it becomes apparent that a major cause of this IBS weakness is the fact thal they pay, on the average, 17.7% more for GM crash parts than their. competitors, the dealer-installers. On the other hand, respondent' statistics on overall growth of IBS open up a dimension which only hinted at in the testimony: the fact that, if exits are common the IBS business, entry is even more so, and that, in the aggrega IBS have shown a significant net growth in the last decade. In short, the record in this proceeding does not establish a cl and direct link between GM's selective distribution policy for cr parts and any weakness of the IBS, as a class. It is clear, at the s' 464 Opinion time, that in general IBS are at a substantial competitive disadvantage with GM dealer-installers, (71)due to the near impossibilty of the IBS overcoming or otherwse compensating for their signifcant price disadvantage on GM crash part. It is also clear that on any local basis the shifting crowd of IBS is generally powerless to shake the stabilzed GM dealer s position in auto body repair work, which is buttressed by favorable prices on crash part. Overall, we believe the IBS are competitively injured by GM's distribution system and that legally that injury meets, if barely, the required showing of substantial injury to competition.

4. Substantial Business Justification General Motors argues that there are two reasons to resist change in its selective distribution system for crash part: the system encourages new car sales and the system would be extremely costly to revise. The second argument is particularly persuasive. The relationship of selectively distributed crash part to new car sales is tenuous. We note that GM has not made the argument that its selective distribution for sole source part is designed to ensure that crash part get into the hands of only qualified installers. There are no product quality or safety issues here. Cf. United States v. Bausch Lomb Optical Co., 321 U.S. 707, 72&-729 (1944); Trpoli Co. v. Wella Corp., 425 F. 2d 932 (3rd Cir. 1970). Crash part are not items which require much special knowledge to install, and the record contains no suggestion that the IBS do not do at least as good a job as the dealer-installers. Nor is a crash parts "exclusive required to reward a dealership for expanding its promotional effort in order to stimulate demand for crash part. These items are not the sort of new or complex product for which the consumer has to be wooed into spending discretionary dollars. Cf. Sandura Co. v. J39 F.2d 847 (6th Cir. 1964).

GM argues generally, however, that by sellng crash part xclusively to its dealers it ensures their loyalty to GM and their creased effort in sellng new cars. A common sense evaluation of e relative importance of crash part sales vis-a..uis new car sales in a alership s profit picture support the ALJ's findings that a dealer , suffcient incentive to sell GM's new cars without the extra ucement to loyalty of a crash part monopoly, and that GM' ,h part policies do not add significantly to that incentive. (IDF 390) Moreover, GM produced no testimony or documentary enee to show that it is having trouble attracting competent , and can only persuade them to take on a franchise by adding Iducement of a crash part exclusive. (72) GENERAL MOTORS CORp.

Opinion 99 F.

On the other hand, GM is probably correct in its claim that a consumer s decision to buy a new car, particularly on repeat purchases, is influenced by the consumer s perception that crash part wil be widely available should the car ever need body repair work.

Availabilty must be distinguished from cost. Consumers look to their insurers to pay accident repair costs (CX 7815P) and rarely, if ever, inquire about cost of crash part when negotiating the purchase of a new car. (See, e. the testimony of dealers Bogard, Tr. 10564-5; Durbin, Tr. 1382; Shaeffer, Tr. 10729-30) Consequently the wholesale cost disparity between dealer-installers and IBS is probably unknown to consumers. In determining which car a consumer selects, such factors as car styling, car price, gas mileage car resale value and even the reliabiliy of engine part assume a greater importance than the cost of crash part. (Sutliff Tr. 1106B- 71) Weare convinced that most consumers are unaware of their dealer s priority rights to crash part, and so few, if any, new car buying decisions are influenced by that aspect of GM's crash part distribution system.

However, whether GM's distribution system assures general availabilty of crash part does concern the consumer. And here, at first glance, it seems that GM could only benefit by getting crash part into the hands of IBS as quickly as possible and at nondiscriminatory costs. The IBS could handle a large percentage of GM vehicle crash repairs. If there is any validity to the argument that unavailable crash parts would cause a car owner to switch to another make of vehicle the next time, then it is reasonable to conclude that the same buyer will be unhappy if she or he experiences delay in getting the car repaired by an IBS.

This leads us to GM's second line of argument. It concerns the risks of change. In our analysis of this question we have relied considerably upon the testimony of Paul R. Murphy, comptroller and principal financial offcer of General Motors Parts Division (GMPD). Starting with the proposition that GM's goals are to distribute crash parts effciently and to gain consumer goodwill with easy availabilty of crash part, Mr. Murphy emphasizes that "the system works" in achieving those goals. (T. 10069, 10071) Although GM' s distributior system for crash parts is selective, it extends to a nationwide (73 network of dealerships, and would produce a reasonable degree c availabilty even if none of the dealers wholesaled crash part on th side. (IDF 116; Murphy Tr. 10077) Availabilty is of course improv, when dealers wholesale, and GM from past experience can expe , 464 Opinion between 20% and 40% of its dealerships to redistribute crash part.

As we noted before, we do not think that this level of availability is the highest possible: inclusion of IBS in the direct GM distribution system might improve it. However, there would be significant costs associated with opening up the system.

GM' s present customers for crash part, the dealers, are also customers for the other 265 000 sole source parts which GMPD distributes. (IDF 159, 167) These other parts are not much described in the record but appear to include .chassis part, interior trim parts and engine part, most of which, like crash parts, are uniquely configured to the various GM cars. (IDF 165; Murphy Tr. 10177-81) GMPD makes no distinction between crash parts and all other parts required for servicing a GM vehicle. (IDF 167) When GMPD was set up to provide centralizd part supply to the dealers, inclusion of crash parts within its distribution system was both logical and effcient. Also, the fact that these crash parts customers have an ongoing business relationship with GM for the purchase of parts, not to mention new cars, means that GM has sigificant leverage in exacting prompt payment for crash parts. GM would not have that leverage with the IBS. If the system were opened, GM would undoubtedly face a significant administrative burden in checking the creditworthiness of and attempting to collect payments from a potentially vast number (74Jof new customers. Other costs could be expected, chiefly those associated with handling more and smaller orders. Not only are the IBS generally smaller operations than dealerships, which leads us to expect smaller order size from them but the size of dealership orders would also shrink, as dealers let GM take back all or a portion of their wholesaling function. When the order size decreases, the cost per ordered item increases. As Mr. VIurphy put it just almost every thing in the entire process is going o . . . require more resources." (Murphy Tr. 10079; Bentson Tr. .. GM h8B no reord from which one ca compute precisly how many GM dealers wholeae part, much les man routinely 8Wck part in amounts greater tha their own boy shop nee and stiU les how many of Be part are crash part. (Murphy Tr. 10293) GMPD oomptroller Paul Murphy eatima.te that 40% of GM' lera are "active" wholeaeni. (Murphy Tr. 10074) However, a6 Mr. Murphy did not quantify what he meant by ve wholesing, and !l an internal GM document indicate that it taes a yearly $50 00 part volume to tge in sigifcat wholesing (CX 725), we agee with both complaint oounal'a and reapondent's expert es that Ii yearly volume of $100 00 or more in part !lies indicate meaningful wholeoling effort. (NeiBOfi 5148; Benaton Th. 16104) Using the $100,00 annual saes cut off point, in 1976 approximately 22% of all GM rs accunte for approximately 80% of all dealer wholese saes of part. (RX 30) GM hf no data, nor doe rd conta data from other source which shows how many deaets atok crasb. part, or tb.e degree to they Btok. (Murphy Th. 1029) , GENERAL MOTORS CORP.

Opinion 99 F.

15811, 15829-30) Unfortunately, the only estimate of the extent of these added distribution costs seems to us to be inflated. Nevertheless, we recognize that such costs would not be insignificant and to some degree would occur on a continuing basis. We also note that other car manufacturers, who presumably have the same incentives as GM for finding the most efficient crash parts distribution system, all use systems similar to GM' s. (IDF 94) This situation, therefore, presents a.more diffcult balance than we faced in Donnelley. In Donnelley we stated In examining the question of business justifications, the economic self interest of the monopolist would be the major but not the exclusive consideration. Where there is little justification for a business policy, the antitrust laws can require that the monopolist take into account the effect on competition of its actions in the line of commerce made up of (those wishing to deal with it." 95 F. C. at 82. In that case respondent offered "no explanation whatsoever" for its refusal to deal. Id. Moreover, it knew that rescinding its refusal (to list commuter connecting flghts) would cost only $6000, and when the actual change in policy occurred it was accomplished "with apparent ease and no il effects. Id. Commissioner Pertchuk seems to suggest that since transaction costs wil always be associated with imposition of a duty to deal, they should be given small weight as a business justification. We disagee. De minimis or speculative transactions costs wil be given small weight, but reasonable projections of significant changeover costs must be heeded. In this case GM has shown that a course of selective dealing gives it satisfactory market penetration with as lean a distribution system as possible. It has also shown that sizeable costs would accompany an expansion of the system, without any (75joffsetting gain to GM."' Certainly if systemwide distribution costs were to increase; consumers would soon face higher repair costs on crash-damaged vehicles. Finally, we are concerned that any order restructuring GM' distribution system would involve the Commission in ongoing supervision of the system. For example, since we found injury to competition only at the installation level of competition, our order would not require that GM deal with all potential customers, but merely deal on equal terms with all crash parts installers. However . Mr. Murphy, without elabora.tion, state it would cot GMPD $40 milion more annually to handle an additional 10,00 customern. ('. 1009) This a.mounts to $4 per customer, which sems hih, given that GMPD' s warehous system is fully computeri. (M\uphy Tr. 10097). .. It must be remembere that, owi to the lltatic demand for cf8!h part, II change in distribution system will not sigificantly incre the actual number of part GM sella in any year. Of rours there cOllldbe 8. shift j" inventory from the deaeI' who are now wholeeing back to GMPD , and GM could rai the warehous price t crash part to cover ita increa wholesing costa, but thi wouldbe R reaHoction of exiting expeno and profi rather th8 any re gain to GM.

610 FEDERA TRE COMMISSION DECISIONS 464 Opinion it would be diffcult to arrive at a clear definition of "crash part installer" for purposes of the order: would it include an IW who opened a small boy shop adjacent to its warehouse? We foresee that we would have to commit extensive resources to reviewig GM' int;rpretations of to. whom and at what price it should sell crash part.

In sum, although there is a troubling degree of injury to competition shown in this case, a reasonable degree of business rationale for the situation has also been shown. On balance, while we are not convinced by this record that GM would he burdened to the extent it has argued by opening its warehouses to IBS, we cannot say that its refusal to do so is arbitrary and without business justification.

III. PROCEDURAL ISSUES A. Noerr-Pennington Having decided that GM and NADA did not engage in a combination to restrain trade, we need not reach the question, raised by respondent in its appeal, of whether the Noerr-Pennington doctrine protets the effort of GM or its dealers to get the FTC to settle the crash part controversy. (76) B. Struck Testimony By "Order Modifying Order Granting Motion of General Motors Corporation for Production of Intervew Report" dated October 31 1978, the ALJ ordered the testimony of four GM witnesses struck from the record, because GM's counsel refused to give complaint counsel written report of pre-trial intervews with Arthur H. Cann part manager of Courtsy Chevrolet in San Jose, California; Joan Mack, part manager of Tom Parsell Chevrolet in Charleston, South Carolina; David W. Vulbrook of the National Asociation of Independent Insurers from DesPlaines, Ilinois; and Henry Faulkner, an Oldsmobile dealer from Philadelphia, Pennsylvania. These report "ere all made by GM attorneys. GM objects to the striking of the testimony on three grounds. First, GM claims that the obligation to urn over intervew report derives from the Jencks Act, which nposes such an obligation on complaint counsel but not on a ,spondent. Secondly, GM argues that absent such obligation, a 'Spondent' s intervew report are "attorney work product" which e privileged from disclosure unless some extraordinary need is own. Such extraordinary need, GM claims, has not been estab- 'led. Lastly, GM states that the struck testimony was "materially GENERAL MOTORS CORP.

Opinion 99 F.

responsive to the assertions of complaint counsel in this proceeding and its striking was, therefore, prejudicial to respondent. (RAB 27- 30) Complaint counsel reply first, that substantial authority as well as fairness compels the adoption of a policy which requires both sides to produce relevant, previously recorded statements of witnesses. Secondly, complaint counsel argue that these report are verbatim witness statements and so are not shielded by the attorney work product doctrine. Finally, complaint counsel deny that GM has suffered any prejudicial injury from the ALJ' s action, and argue that even if it did, such injury would be entirely proper in order to prevent a partial view of issues. (CAAB 59) 1. Procedural History Respondent startd an avalanche of motions and orders on this subject by a small snowball, requesting copies of complaint counsel's witnesses, to be delivered one week before the witnesses were scheduled to testify. In response to this motion the ALJ ordered both sides to exchange witness interview report in advance of scheduled testimony. (Order Granting Motion of General Motors Corporation for Production of Interview Report, April 10, 1978, hereinafter Order of April 10, 1978). Complaint counsel moved for reconsideration, on the grounds that interview report are producable only after testimony, but the ALJ denied this motion and reaffrmed the Order of April 10, 1978. Making the best of, in their opinion, a bad business, complaint counsel then moved for production of respondent' s intervew report. Counsel for GM replied that he had nothing in his fies that was not exempted from disclosure by either the attorney-client priviledge or the attorney work product doctrine. (77) Matters Came to a head at a hearing on September 27, 1978, where the ALJ reviewed report respondent was withholding, which related intervews with the four witnesses named above. After literally blue pencilng out certain portions which he deemed attorney work product, the ALJ ordered that the remainder of the report be produced for complaint counsel forthwith. This counsel f01 GM continued to decline to do. Complaint counsel thereafter move, to strike the testimony of the four witnesses, although subsequent! entered into a stipulation that the testimony of Cann, Mark aI Faulkner could stand. (It appears that complaint counsel we actually present at respondent's counsel's intervews with th. witnesses.) The ALJ decided, nevertheless, that the better cou would be to apply the sanction to all four witnesses, and so orde on October 31, 1978.

464 Opinion Applicabilty of Jencks Act Procedures A "Jencks Act" statement is defined, 18 V. C. 3500(e), as a written pretrial statement made by a witness for the united States and signed or otherwise adopted or approved by the witness; or a stenographic, mechanical, electrical, or other recording, or a transcription thereof, which is a substantially verbatim recital of an oral pretrial statement made by such a witness and recorded contemporaneously with the making of the oral statement. (Statements made to grand juries are also included but not relevant here.) The Jencks Act proper applies only to criminal prosecutions. We have held, however as a matter of our discretion, that its principle should be applied in Commission proceedings to require production of certain prior statements by complaint counsel's witnesses after they have testified. See US Life Credit Corp., 91 F. C. 984, 1037 (1978), and cases cited therein. Complaint counsel now urge that we apply the Jencks Act principle to respondent' witnesses to uphold the ALJ' s action in striking the testimony at issue here. We decline to do so for the reasons set forth below.

At the outset, we wish to make clear to both counsel and the administrative law judges that our reliance on the Jencks Act principle" rather than the Act itself (whose limitation to criminal cases we have already noted) is not an invitation to ignore the other salutary limitations incorporated in the Act to protect the principle integrity. In particular, we reiterate that (1) the safeguards of the accuracy of the statement (i. the requirement that the witness have unequivocally (78Jadopted it or that it be substantially verbatim and complete) must be strictly observed US Life, supra; and (2) production will be required only after the witness has testified fnterstate Builders, Inc. 69 F. C. 1152, 1165-7 (1966). So far as appears from the record before us, neither of these mitations was honored in the series of rulings which led up to the riking of the testimony at issue." However, although the point is . As we indic8te in Interatate Buildrs thi procedure is desiged to protect against a chiling effec on the ingnes of peralU with information usful to Jaw enforcement to make statementa in the cours of an tjgation although they might not be willing to testify at trial, and considern the limited purp of the Jencks , enabling the defendant/respondent to make UB of 1. prior inconsistent statement in cunducting CtQ intion. The Act ha never ben intended 8B a tol of disvery, since a respondent iBfree to conduct hib own ':ews of potentia or planned witness.

Se our diBUBion in US Life 91 F. C. at 10389, of the riorous teta which must be applied to qualify ents either lladopte by the witness or 8B suootatially verbatim recita, in particular our obarvation from Palermo v. Unite Srota 36 U.S. 343, 352 (1959), that "the leglation W8B designed toeliminate the of distortion and misreprestation inherent in 1.report which merely selec portions, albeit accurately, engthyoralreita.

GENERAL MOTORS CORP.

Opinion 99 F.

not entirely free of ambiguity.' it appears that the ALJ disclaimed reliance on or limitation by the Jencks principle in the rulings under consideration." In (79)these circumstances we need not and do not decide whether the application of the Jencks Act principle can or should include prior statements by respondent' s witnesses as well as those of complaint counsel." What we must decide is whether the ALJ' s order has some suffcient basis other than the Jencks Act principle.

Attorney Work Product As we said in Interstaie Builders, supra 69 F. C. at 1172, "(i)n view of the limited nature of the Jencks rule, it is clear that the policy considerations underlying the work product rule which were so emphatically stressed by the Supreme Court in Hickman v. Taylor (329 U.S. 495 (1947)) are stil operative whenever Jencks statements are not involved.

In the seminal case of Hickman v. Taylor the Supreme Court granted qualified immunity from discovery to an attorney s work product described as private memoranda, written statements of witnesses, and mental impressions or personal recollections prepared by an attorney in contemplation of possible litigation. From the onset, the Supreme Court, and the Commission likewise, have recognized that the work product rule is not absolute but may yield on a showing of substantial need. Hickman v. Taylor, supra 329 U. at 509; see, e. , Warner-Lambert Company, 83 F. C. 485 (1973); Graber Manufacturing Company, Inc. 68 F. C. 1235 (1965). In both C. cases, the Commission emphasized that before respondent could receive documents considered to be the attorney s work .. The AL'II Order Denying Motion For Re1lideration Of Order Granting Motion of General Motors coplaint counslQlrpration For Pruction of Intervew Report, April 20,1978, Ilpe in term of reui hand over Jencks IIUttementa," and involved the AL' s initial decision on this subjec USin Life (liubauently overtumed in thi repre by the Commision in the cite above). n In hi Order Denying Motion For Rensideration Of Order Of September 21, 1978, Reuiring the Pruction of Intervew Report, date Ocber 13, 1978, the AI IWrt that "my Order ia neither precate on the Jencks Act nor on the Commifion 8 expreionu ita adoption of the principles inherent in Jencks. rerd He added that whie atatementa covered by the Act would clearly be included in the larger cl8B of 8tatementl lIubject to the order, that order "nee not be and was not intended to be limited to Jenck Act principlet. U In US Life, aupra 91 F. C. a.t 1037 n. 34, we obsrv that we had not theretofore ben caled upon to rt$lve that issue, simply notin the Supreme Court s decision. United States II. Noble, 422 U.S. 225 (1975). Complaint count urged reliace on that ca here. We obsrve, however, that in that criminal proution thf proof substatially consiste of two eyewitnes identifcations of Noblet, that the sttements at iaue whid purprty undermined the witness' creibilty had ben taken by 8 defena inve6tiator whom defense couns sought to put on the stand to testify to the IIUbsce of the atatementa, and that thedecison !lms to have tume on the propoition that thi attmpt waived any work-product privilege which attached to the atatements. 422 U. at 23240.

. .

464 Opinion product, strong showing of special circumstances, good cause or necessity must be demonstrated. BB (80) The work-product rule has now been formally incorporated in the Commission s rules governing discovery, at Section 3.31(b)(3). Although the Commission rule did not become effective until after the issuance of the rulings in question, it is nonetheless expressive of the work-product rule as derived from Hickman and as expressed since 1970 at Rule 26(bX3) of the Federal Rules of Civil Procedure. As such it embodies the principles appropriate for application to the present question. Rule 3.31(b)(3) provides in pertinent part that: a party may obtain discovery of documents. otherw disoverable under subdivision (bXl) of the rule and prepared in anticipation of litigation or for hearing by or for another party or by or for that other party s representative (including his attorney, consultant, or agent) only upon a showig that the party seeking discovery has substantial nee of the materials in the preparation of his ca and that he is unable without undue hardship to obta the substatial equivalent of the materials by other means. In ordering discvery of such materials when the required showig has ben made, the Administrative Law Judge shall protect against disclosure of the mental impressions, conclusions. opinions, or legal theories of an attorney or other representative of a party.

First, there seems to be no question that the witness statements or intervew report or attorneys' notes in question are within the class of documents "prepared in anticipation of litigation" and covered by this section and the case-law doctrine. Indeed, the Hickman decision itself concerned just such pretrial statements or reports. Second, it must be noted that the rule (and the doctrine) apply what might be characterized as two layers of protection. Any document which falls within the coverage of the section is to be ordered produced only upon a showing of "substantial need" and "undue hardship. " It is only when that showing has been made that the further admonition to "protect against disclosure of the mental impressions, conclusions opinions, or legal theories of an attorney or other representative of a party" comes into play. Thus, the task of an administrative law judge faced with a request to compel production of work product or hearing preparation materials consists of more than simply excising menta mpressions, etc., and turning over the residue. The ALJ must first consider whether substantial need for the materials has been emonstrated, which need cannot be substantially met from other IUrces without undue hardship. See Order Upon Application for lterlocutory Review In re The Gillette Co. FTC Docket No. 9152 ecember 1, 1981) (98 F. C. 875). (81) The fac that both Wanwr-Lambert Company and Grober conoorned the work product of complaint counsel not diiniBh their prentia value in view of our endorsment of the priciple of even-hadedes in very. AllBeou Industri 72 F. C. 1020 (1967). GENERAL MOTORS CORP. OJOJ Disnting. Statement 99 F. In the present case, this analysis does not appear to have been carried out. Rather, the ALJ essentially grounded his rulings on his view that Commission precedent requires that "each side should maximally provide the other with information as to the positions each wil take and the evidence each will use to prove its case. Order of October 13, 1978 supra. While we have no quarrel with this statement as a characteriztion of the policy underlying. our discovery rules and practice, it must obviously be qualified in practical application by reference to those rules and to our rulings in such cases as US Life and Interstate Builders. Further, we note that the relationship between this stated principle and the specific disclosure ordered by the ALJ here is not clearcut; that is, disclosure of witnesses' prior statements is not especially well-targeted to reveal a party s "position" or "the evidence (it) wil use to prove its case." Other devices, far less intrusive and better aimed, are available to serve these purposes. The only end which disclosure of such statements can uniquely serve is that of cros-examination, and only then when the safeguards of the Jencks Act principle are observed.

We are constrained to conclude that the ALJ' s disclosure rulings upon which he based his order to strike testimony were not properly grounded either on the Jencks Act principle or on a due consideration of the work-product doctrine. Weare not of a mind to search the record in an attempt to determine if the orders might have been justifable upon proper analysis, particularly in view of the fact that complaint counsel themselves stipulated that the testimony of three of the four witnesses need not be struck, and gave no compellng reason why they needed to see intervew report for the fourth. Under the circumstances, we believe that the ends of justice are best served by reinstating the testimony of the four witnesses and considering it in reaching our decision on the merits, and we have done so. Thus, respondent has not been prejudiced by the ALJ' Order of October 31, 1978, or the series of orders which led up to it. DISSENTING STATEMENT OF COMMISSIONER PERTSCHUK I dissent from my fellow Commisioners' decision to dismiss the complaint in this matter. I would have found liabilty on the par of General Motors for a violation of Section 5 and would have voted to issue an order requiring GM to deal on non-discriminatory terms with independent body shops.

. Any order to GM to dea on a non-driinatory basis would have ilinclude cert with the maority that GM ha no duty to dea with independent wholeseradiretlysinceit ha choon toag qualifcatiolU. Firat, I8aume the "firs tier wholes function itslf and there is no duty on patheof.l rimnopolist ttder the tbeo (Qmtinur 616 FEDERAL TRADE COMMSSION DECISIONS 464 Disnting Statement As I understand the majority s opinion, the Commission does not retreat from the position it took in Reuben Donnelley, where we stated that a monopolist has a duty to deal fairly under certain circumstances with those seeking to deal with it. Nor does the Commission fail to find adequate harm to competition to constitute a violation. To the contrary, the majority concludes that there is " troubling degree of injury to competition shown in this case. (Majority op. at 75) The Commission chooses to dismiss the complaint, however, on the basis that there is adequate business justification for GM's refusal to deal with independent body shops. I disagee with this conclusion. (2) In assessing GM's argued justification, it is useful to review the standard set out in Reuben Donnelley for an adequate business justification for refusing to deal with certin classes of customers. There the Commission stated that a duty to deal arises only if a failure to do so results "in a substantial injury to competition and lacks substantial business justification.'" (emphasis added) Thus, we are not concerned with only minimal or speculative harm to competition, but, in the event that substantial harm is found, the justification, similarly, must not be minimal or speculative, but must be substantial. Moreover, the burden of proof on this issue is the respondent' The injury to competition from GM's refusal to sell crash parts to independent body shops is that these suppliers of repair servces are disadvantaged in competing with GM dealers because they must pay substantially greater costs for part than their dealer competitors. A part of this differential is theoretically accounted for by the second level "wholesaling " function of dealer rdering and storing part. The only record evidence of the costs of ordering part is that it constitutes 2% of the dealers' sales price. (Majority op. at 52 , fn. 67) The record is silent on the percentage of dealers who store parts in large quantities, either for their own use or for resale to others. Many dealers, who receive the maximum discount, apparently order crash parts only on an Uas needed" basis, however. The "premium" paid by independents, that is, the price paid for crash parts over and above the price paid by dealers, amounted to $41.6 milion in 1977. If the 2% cost attributed to the wholesaling If Reuben Donnelle 00 crete rival to it6lf (Mority op. at 43) In addition, there are legtimate rena for GM to ('fus to dea with an CWitomer under particular circtUt8ce, wcluding hih riks of granting creit, por ayment hitories, etc., which would have to be acknowleded exceptiona to a duty to deal requirement. . The Reuben H. Donrudky Corp, 9f F. C. 1 d sub rwln FT v. w.IAirline Guid 63 2d 920 'd Cir. 1980), cert. denie 101 S. Gt 136 (1981)(45 U.S. 917 (1981)J. . 95F. at82.

. It i6, of OOUI", dicult to estiate .agegat. exce price pad by the disvantaed dealers frol1 the xard. In 1977 aM pad approxhilately $98.8 million to dealers for CTllh part under the wholese compensation (ContinUf) GENERAL MOTORS CORP.

Disenting Statement 99 F. T. function is subtracted, this premium, caused by the refusal of GM to deal directly, amounted to $36.9 milion in 1977 alone. (3) Moreover, this price differential is not the only advantage dealers have from GM's system. For every part GM sells to a dealer who in turn resells it at wholesale to an independent, GM pays the dealer 30% of the wholesale list price of the part under GM's wholesale compensation plan. While this system no doubt encourages dealers to resell part-the reason it was urged by the Comission at its inception in 1968-it is also costly to GM. In addition, there is a not insignificant amount of uerroneous claims" and abuse accompanying the system, including claims for wholesale compensation by dealers who sell to other GM dealers. (IDF 111-112) Against this competitive harm to independents and the losses to GM from the wholesale compensation plan, GM offers two justifications for its refusal to deal with independent body shops: 1) the system encourages new car sales by acting as an inducement to dealers to affiiate with GM initially and to sell additional new cars; and 2) opening up the distribution system to independents would be costly and entail some uncertainty. The majority specifically rejects the argument that incentives to dealers are significant to the degree necessary to constitute adequate justification for refusing to deal with independents. (Majority op. at 71-72) The majority, however does accept the second argument as adequate justification. We should be wary of justifying a monopolist' s refusal to deal with a class of customers where competition is harmed substantially on the grounds that there are .transaction costs in dealing with additional customers. A challenge to a monopolist's refusal to deal by its nature, involves refusing to deal with a class of customers. Consequently, it wil always be possible to posit that the transaction costs of dealing with additional customers justifies selective distribution. While it is certainly true that a more extensive distribution system is more expensive to manage, and should not be ignored in assessing net competitive effects, our focus must be on whether it outweighs substantial harm to competition. The only record estimate of the extent of these added distribution costs was provided by Mr. Murphy, who estimated that it would cost $40 milion to deal with 10 000 additional customers, or $4 000 per customer. (Tr. 10079) The majority concedes this estimate is "inflatplan. (T. 16277-8) Th amounts to 30% of the dealer price of qualified Bales to wholesale customers. (IF 73 Consuently, the agate dealer (or wholealer list) price for part sold to wholesale customers WB approria.t.ly $327 milion and the actual amount pad by thes C\Jtomerf was about $235 million, BIuming the received a 28% off lit diunt. (Se majority ap. at 49) Consuently, if the dealer price ia reduce 17.7% from tt independents, the tota premium pad is about $41.6 milion. 464 Concurr Statement ed." (Majority op. at 74) Moreover, the testimony was given without explanation as to the components of the costs, nor without a clear explanation of whether the costs are one-time start up costs or recurring. Against this apparently (4Jrather speculative estimate, which the majority concedes is inflated, the record showed excess prices of almost $37 milion in 1977 on an annual basis. I cannot agee that GM has carried its burden in showing that the additional transaction costs of dealing with independent body shops outweighs the substantial harm to competition shown by complaint counsel. The record falls far short of proof adequate to justify a system which clearly raises prices to independents by milions of dollars and which, no doubt, drives up significantly the costs we all pay for dente fenders and crushed bumpers. For these reasons dissent from the dismis of the complaint. CONCURRING STATEMENT OF COMMISSIONER CLANTON I concur in the Commission s decision in this matter, but I have some observations about the rationale used to reach the legal standard in Part II.C. of the opinion. That standard, which is articulate on pp. 36 and 39, is identical to the approach adopted in our decision in The Reuben H. Donnelley Corp. 95 F. C. 1 rev d sub nom. Federal Trde Commission v. Official Airline Guides, Inc., 630 2d 920 (2d Cir. 1980), cert. denied 101 S.Ct. 1362 (1981)(450 U. 917 (1981)). Nevertheless, while applying the substance of the Donnelley standard, the Commission s opinion in this case suggests that the precedent of Donnelley is unnecessary to the decision here since there is an adequate independent basis in prior cas law for performing the same analysis and reaching the same result Associated Press V. United States, 326 U.S. 1 (1945); United States V. (1912); Grand CaillouTerminal Railroad Assn. 224 U.S. 383 Packing Co. 65 F. C. 799 aff'd sub nom. La Peyre V. Federal Trade Commission, 366 F.2d 117 (5th Cir. 1966). at least in Presumably, the Commision s analysis is designed, s ruling inpart, to avoid direct conflct with the Second Circuit' ')fficial Airline Guids. I have two problems with that approach. In he first place, applying the Donnelley theory without calling it by hat name is highly unlikely to change the outcome of judicial . If indepeden were allowe to bllY GM cra pa direly from. GM. we would expe to fi not only 8 but an 1note in volume done by them as people luction in price cha by indapendentI, due to lower cot!, .k advantage of lower prioe, and a decre in price charg by the deael" as they reponded to the tlpetition. Con&uently. the tota spent by consumenl could be reduce by more tha the exce wnount for 9h part pad by independenta. We should kep in mid tht GM's we of crwh is no small enterpri. "lplaint ooUlI eeimate that the dollar volue of domestic shipments of thes pa Wil about $59 milion in 7. (Maorty op. at 6) GENERAL MOTORS CORP. 619 Final Order 99 F.

review in future cases. Certinly, such an approach wil not succeed in the Second Circuit. Whether another court of appeals wil agree with our Donnelley analysis is unknown ' but it is doubtful that the manner in which we characterize the rule wil influence the end result. My second concern is that the analysis in this opinion seems to suggest a closer link between prior precedent and the Commission s present legal theory than we acknowledged in Donnelley. Donnelley, the Commission recognized that the imposition of an obligation on lawful monopolists to deal fairly (or not arbitrarily) with firms seeking access to the monopolist' s products or services fell outside the mainstream of monopolization law. Although in Donnelley the Commission felt that it was a "small step" from analogous case law to the standard enunciate (2)in that case, it was clear that the rule developed there did involve an extension of existing law, or at least the first clearly exclusive reliance on such an approach by either agency or court.' I ageed with that analysis and continue to do so. To the extent, however, that the Commission s opinion here suggests a different reading of the legal duty imposed upon no,,colluding monopolists by the cases relied upon in Donnelley, I would disagee.

FINAL ORDER This matter has been heard by the Commission upon the appeals of complaint counsel and respondent from the initial decision and upon briefs and oral argument in support of and in opposition to the appeals. For the reasons stated in the accompanying Opinion, the Commission has determined to sustain respondent's appeal. Complaint counsel's appeal is denied. Accordingly, It is ordered That the complaint is dismissed. Chairman Miler did not participate.

Commissioner Pertchuk dissente.

, It is pDible that the Fifh Circuit Court of Appe might reach I! conclWlion different from that of theSend Circuit. Se Fulton v. Hecht 68 F.2d 1243 (5th Cir. 1978); La /'re v. Fedrnl Tr Commision, 36 F. 117 (5th Cir. 196).

. While lanage in La Pere de6ribe a "duty of a lawful monopolist to conduct ita businee in such a way 11 to avoid inflcting competitive inury on a c1tU of cuatomar8 " 36 F.2d at 120, that dictum did not describe the ilue squarely before the court and it WB! not necry to ita holdi. GODFREY CO. 6:l1 621 Modifying Order