advantage. This learning process is said to be the key to invigoration of the U.S. industry. Although these objectives are legitimate, even commendable, I believe that a joint venture between GM and Toyota is a plainly anti-competitive strategy for achieving them. Furthermore, I believe that GM's objectives can be largely realized through competitively innocuous alternatives. My basic argument can be represented in five points. 1 First and foremost, I believe that in the absence of the joint venture with Toyota, GM would undertake, hasten or expand a contingency plan for U.S. assembly of a vehicle originating with its Japanese affiliate, Isuzu. These so-called "R-cars" are the same vehicles that GM hoped to import in considerable numbers as completely-built-up units for distribution through its dealership system. While such an undertaking is not equivalent in all respects to the Toyota joint venture, it clearly raises none of the competitive dangers posed by association with Toyota. Isuzu is a very small factor in the U.S. or world markets, it is already more than one-third owned by GM, and there can be no question of anti-competitive coordination resulting from a GM-Isuzu production operation. It is, I believe, clearly to be preferred on competitive grounds to a GM-Toyota linkage. Let me make clear that this is not an alternative which I conjured up, but 2 rather one for which there is much support in the factual record. Let me This contention is contained in the expurgated version of my memo released by the FTC, on page 424 of that release (page 66 of my memo), hereafter cited as "FTC p. 424 (Kwoka Memo, p. 66)," and "Dissenting Statement of Commissioner Patricia P. Bailey," December 22, 1983, p. 12. 2See FTC p. 108 (Bureau of Competition Staff Memo, ch. II, p. 41). further note that either a Toyota or Isuzu association would involve new production capacity by GM, since I am convinced that their small car needs require new sources of vehicles. But the joint venture with Toyota raises clear and substantial risks of coordination. That coordination, in my view, is likely to result in higher prices supported by diminished GM production of its other small cars (Chevette, etc.), relative to the Isuzu alternative. Secondly, I believe Toyota's output incentives in this joint venture are mixed. It will profit through some combination of joint venture operations plus the sale of major components. But Toyota has ambiguous incentives regarding U.S. production, since higher output here lowers small car prices and consequently the profits it earns on imports from Japan. Indeed, in the presence of the Voluntary Restraint Agreement (VRA), Toyota presumably would like to see U.S. production reduced, so as to maximize returns on its import quota. The VRA was intended to aid U.S. companies, but it has also succeeded in producing "exceptionally lucrative profits" for the large Japanese importers and has led 3 them to resist expansion or entry by the smaller Japanese companies. Although joint venture profits moderate this incentive to Toyota to some degree, the joint venture simultaneously gives it some measure of control over GM's small car sales in the U.S. This occurs because the joint venture has the properties of a supply arrangement, with GM becoming dependent on Toyota for an important part of its product line-up. Such control is competitively worrisome, because GM's sales are a decisive element in the price Toyota in turn obtains on its imports into the U.S. 3FTC 312 (Koch Memo, p. 4 ). Regarding smaller Japanese companies, see "Japan Seen Accepting U.S. Compromise on 'Voluntary' Auto-Export Limits," Washington Post, October 26, 1983, p. A23. Third, I believe the joint venture is certain to facilitate anti-competitive information exchanges between GM and Toyota. Clearly some information flow is even partly its purpose, one might argue. But necessary for the joint venture the line between the permissable and impermissable is not clear, and the parties' private incentives are not identical with the best interests of the market. The companies will be brought into close contact, have more consonant objectives, and benefit from vastly improved channels of communication through this joint venture. Here too I am not relying on abstract possibilities which I conjured up. As I stated in my memo, anticompetitive information exchanges probably have already transpired," involving sensitive product information and retail prices 5 differentials. Apart from the coordination these may already have facilitated, such exchanges are indicative of precisely the sort of problems inherent in this joint venture. Arguments to the contrary that these exchanges were necessary for the joint venture, that the joint venture does not facilitate whatever information the companies might want to exchange, or that the companies would not reveal competitively sensitive information to each other are in my view belied by the evidence already in the record. Fourth is a related concern over the transfer pricing formula. Given the supply aspects of the joint venture, some mechanism for pricing the product is necessary. Periodic discussion between GM and Toyota would probably be the most worrisome from a competitive point of a view, since that is most likely to reveal general pricing strategies and to spill over to a wider range of impermissable topics. At the other extreme would be some device over which the parties have "FTC p. 455 (Kwoka Memo, p. 96). 5"Dissenting Statement of Commissioner Patricia P. Bailey," December 22, 1983, PP. 9-10. See also FTC p. 252-254 (BC Staff Memo, ch. VIII, pp. 17-19). no control. The transfer pricing provision of the joint venture agreement falls somewhere in between these extremes. Toyota Corolla price changes are automatically propogated into the cost of the joint venture vehicle to GM, though not on a dollar-for-dollar basis. Yet the mechanism effectively eliminates the possibility that GM might not follow a Corolla price increase, and thus reduces Toyota's uncertainty about its major rival's response. Further, if other small car producers follow Toyota's lead (not an unlikely occurrance, based on past behavior), the market basket approach of the formula ensures a more nearly matching increase by GM. Apart from these inherent problems, there are two noteworthy procedural dangers. One is that the process of devising the transfer price would seem to require GM and Toyota to negotiate, in advance of the joint venture model year, the prospective wholesale prices of the market basket vehicles, including adjustments to eliminate "the value of equipment changes and product ,,6 improvements. This represents precisely the kind of direct discussion and agreement on future prices and product changes that the "formula" was ostensibly designed to avoid. Secondly, GM and Toyota explicitly reserve the right to reject the computed transfer price whenever it is "at significant variance with then current market conditions," in which case they can negotiate an alternative price. Although one can readily understand the rationale for such an escape-clause, its use makes close analysis of the formula irrelevant. The companies can then discuss prices directly. 6 GM-Toyota "Memorandum of Understanding," February 17, 1983, Exhibit A. See also "Dissenting Statement of Commissioner Pertschuk," December 22, 1983, P. 5. 7"Memorandum of Understanding," pp. 5-6. Fifth and lastly, I have some concerns that this joint venture could weaken Toyota's incentives to enter the U.S. large car market. That market is the traditional source of domination and profitability to the domestic industry, and to GM in particular. GM's grip on the large car market has in fact increased Let me But the over the past few years, as Ford and Chrysler have been beset by a variety of problems. The joint venture with Toyota might signal GM's partial withdrawal from the small car market, to Toyota's advantage, in return for which Toyota might not pursue large car competition against GM as aggressively. hasten to add that I have seen no evidence of any such understanding. business and economc forces for this possibility are in place, and if it should occur, the adverse effects on the domestic large car competition would be huge. There is, in short, a large number of distinct competitive dangers in this joint venture. Collectively, I believe, they are both real and substantial, and make output diminution (or perhaps better stated, a smaller output increase, compared to the alternatives) enormously likely. Both GM and Toyota seek higher U.S. car prices and this joint venture enhances their means of accomplishing this objective. The structure of incentives, the improved means of coordination, and the evidence on the record to me prove the likelihood of such consequences, whether intended or not. Against these concerns must be weighed potential efficiency gains from the joint venture. Few merging companies fail to make claims of efficiencies. Efficiencies are easy to argue, hard to assess, and (the evidence indicates) often do not come to pass. Hence both the FTC and the Department of Justice have established stringent standards for accepting efficiencies as a defense to otherwise anticompetitive structural changes in industries. Conventionally, efficiencies from joint ventures consist of such things as cost savings from improved realization of scale economies, and new products or processes developed by pooling financial resources. The GM-Toyota joint venture |