Pub. 55 2014-2015 Issue 1

31 2014 FALL that new entrants would benefit from a highly vertically integrated structure as they learn and adapt to their new market. As discussed above, Bresnahan & Levin note that there are cases where vertical integration will lead a firm to more efficient outcomes. Our issue is that these firms are all in either the min- ing industry or the transportation/logistics business. We agree that trucking firms and airlines likely benefit from the ability to adjust schedules and remain flexible in light of external factors like weather. However, Bresnahan & Levin offer no opinion on whether vertical integration would solve similar coordination issues in other industries, let alone in automotive retailing. We submit that given the long logistics chains and lead times involved in manufacturing an automobile, any benefits would be virtually non-existent. In this regard, Novak & Stern (2009) 14 offer insights into automobile manufacturing and the extent to which individual procurement choices are interdependent and need to be coordi- nated. They note that manufacturers may internally source clusters of components if they foresee difficulties in coordinating design changes that require large scale modifications. It is unclear to us how this research supports your claim that consumers would benefit from automakers vertically integrating into retailing. 15 D. Restrictions and Past Automobile Retailing Studies You next claim that state restrictions on vertical integration raise prices for consumers without any quality improvements. 16 We would expect such a strong statement to be backed by equally strong empirical arguments and literature. This, however, is not the case. In fact, we note that the FTC studies relating to the automobile industry are outdated and, from our review, appear to suffer from econometric problems. In addition, the other market held out to be the ideal case study for automotive retailing – the retail gasoline market – involves a comparison that we submit is fundamentally flawed, as discussed above. There is a lack of empirical evidence to support any claims that eliminating vertical integration restrictions in auto retailing would benefit consumers. Both Eckard (1985) 17 and Rogers (1986) 18 are outdated studies, and both suffer from econometric problems. 19 Eckard attempted to use disaggregated data to evaluate relevant market area (RMA) laws. Interestingly, Rogers himself criticizes Eckard as having omitted variables, endogeneity problems, and an inadequate model of supply and demand. The Rogers study is no better. Our review of that paper shows serious econometric flaws that resulted from either poor model specification, poor data, or both. Many of the estimated variables in Rogers have the wrong sign or are insignificant, including those of estimated price elasticities, the coefficients for the impact of labor costs and advertising, and the dummy variables for state laws. These are weak empirical citations given their age, their lack of econometric integrity, and the widespread technological changes that have occurred since their publication. To rely on these studies as a basis for considering and deciding major regulatory changes is unwarranted. Lawmakers should consider the clear evidence of high levels of competition that exist because of today’s regula- tory environment to be of much greater importance. It is basic economics that above normal rents are not sustainable in a com- petitive market unless there are barriers to entry. Our industry is open to competition, and there is no evidence to suggest that there are any material barriers to entry in automotive retailing. It is true that new entrants in automotive retailing must invest in showrooms, repair facilities, and inventory. Nonetheless, new individual dealerships are established all the time. Moreover, other retail sectors have effectively the same requirements of working capital, inventory, land, and facilities. New entrants over the past 50 years have built up franchises and gained market share all relative to incumbents. * * * * * There is general agreement that it is possible in some cases for vertical integration to lead to cost savings, even economically material savings. Unfortunately, there is no evidence that verti- cal integration in retailing, let alone automotive retailing, would provide any benefits to consumers. The lack of empirical evidence of benefits, the inability to find any industry parallels, and the high probability of negative unforeseen consequences leads 20 to the conclusion that there is, in fact, no benefit to consumers in ending current restrictions on automobile manufacturers from selling directly to the public. 14 S. Novak and S. Stern, Complementarity Among Vertical Integration Decisions: Evidence from Automobile Product Development, Management Science, 55(2), 311-332, 2009. 15 The final source to which you cite, Forbes & Lederman (2009), studies the relationship between major U.S. airlines and their regional affiliates. Once again we can see no clear parallels to the automobile industry, and none are mentioned by the authors. 16 Your letter to the New Jersey legislature states, at page 6, that “past studies by both academic researchers and FTC staff have concluded that state-imposed restrictions on automobile manufacturers’ ability to negotiate with their dealers increased the prices paid by consumers without leading to notable improvements in service quality.” 17 E. W. Eckard, Jr., The Effects of State Automobile Dealer Entry Regulation on New Car Prices, Economic Inquiry, Volume XXIV, 1985. 18 Robert P. Rogers, The Effect of State Entry Regulation on Retail Automobile Markets, Federal Trade Commission Bureau of Economics, 1986. 19 We recognize that there is a third study (R.L. Smith (1982)) that is often cited in this regard and that these three papers – Eckard, Rogers, and Smith – largely form the basis of nearly all criticisms of the retail automotive system. However, since Eckard and Rogers cover fundamentally the same topics as Smith, we chose not to review Smith here. 20 These unintended consequences could include, without limitation, the consolidation of low volume dealerships, loss of employment, decreased consumer convenience, loss of service locations, and increased prices.