Prepared Remarks of Kevin J. Arquit Director Bureau of Competition Federal Trade Commission Before the American Bar Association Section of Antitrust Law "Resale Price Maintenance: Consumers' Friend or Foe?" August 12, 1991 I. Introduction Good afternoon. It is a pleasure to be here today to talk with you about resale price maintenance. RPM is back in the spotlight. The Commission recently has entered into two RPM consent orders, both of which have received a great deal of attention. Further, the Senate has passed a bill, and a House committee has reported a bill, codifying the per se rule against RPM and modifying the evidentiary requirements for proving an agreement on price in cases involving terminated dealers. The economic effects of RPM remain a subject of lively debate, and opinion is still divided on whether RPM enforcement is in the public interest. Today I will discuss several efficiency and anticompetitive rationales for the use of RPM, as well as the Commission's recent activities in this area. I should note that these remarks represent my views and are not necessarily those of the Commission or any individual Commissioner. Obviously, one reason for our interest in RPM is the fact that the Commission is a law enforcement agency with an obligation to enforce the will of Congress and the courts. Where our staff finds clear evidence of a per se violation we will, of course, recommend to the Commission that enforcement action be taken. We are most interested in acting against egregious violations, such as cases involving express agreements on price throughout a supplier's distribution network. But we are also interested in finding those cases where the practice in question is most likely to harm consumers, because we want to focus our enforcement resources on instances in which consumer injury appears greatest. In addition to affecting the allocation of our resources, the apparent degree of consumer injury may affect the remedy we seek. While this speech is not intended to provide a final resolution to the lively debate about RPM, it may suggest how we approach our enforcement responsibilities. I will start with a brief description of our recent law enforcement actions in the RPM area for those of you who may not be familiar with them. In Kreepy Krauly, the complaint alleged that the respondent, a manufacturer of automatic swimming pool cleaning machines, systematically entered into written agreements with its dealers to maintain retail prices. Our consent order prohibits the firm from engaging in the practice in the future. More recently, the Commission accepted for public comment a consent order with Nintendo of America to settle allegations that the firm fixed the prices at which dealers advertise and sell Nintendo home video game hardware to consumers. The consent order prohibits Nintendo from agreeing with any dealer on the price at which the dealer may advertise or sell any Nintendo products to consumers. The order also prohibits Nintendo from pressuring or coercing any retailer who chooses to discount Nintendo products, requires the firm to tell its dealers in writing that they can sell the products at whatever price they choose, and, for a period of five years, prohibits the firm from terminating any distributor because of that distributor's resale price. The Commission, of course, is not alone in focusing on the RPM area. Recently the Senate passed, and the House Judiciary Committee reported, bills that would ease the showing necessary for plaintiffs to establish an unlawful resale price maintenance agreement. Both bills would codify the existing per se prohibition against minimum RPM, but not the per se prohibition against maximum RPM. The Commission has not taken any formal position on the legislation, and my comments are not intended to take any stance on the legislation. One purpose of the proposed legislation, of course, is to ease the evidentiary burden required to show an unlawful agreement to fix resale prices. Under the Monsanto and Sharp cases, it is clear that a plaintiff must demonstrate an agreement on price or price levels, and that the termination of a dealer who does not comply with a supplier's unilaterally-imposed price policy is not actionable, even if the termination is in response to other dealers' complaints. These cases have fortified the Colgate doctrine, which allows a supplier to "exercise his own independent discretion as to parties with whom he will deal." Some questions remain as to where the Colgate-protected unilateral conduct ends and an agreement on price begins under the current law. For example, how often can a supplier remind its dealers that they might be terminated if they discount? We know from Monsanto that it is not enough to show that a dealer conformed to a supplier's suggested price. Rather, there must be evidence that the dealer communicated its acquiescence or agreement, and that this was sought by the supplier. I would suggest that a supplier is still taking a significant risk if it does more than communicate its price policy to dealers. Any action that seeks to elicit dealer acquiescence in the supplier's pricing policy remains legally suspect. A showing of an agreement can be based on ambiguous evidence. In Monsanto itself, in finding an agreement on price, the Court pointed to evidence that Monsanto officials threatened not to provide adequate supplies to a discounter during the "shipping season" for the product in question." In addition, the Court referred to meetings at which Monsanto officials promised retailers to "ge[t] the market place in order" by ensuring that retailers not "deviate downward" from the suggested retail price." The mechanism cited for the discouragement of price deviations was the "risk of being deleted." This suggests that a supplier can threaten to terminate price cutters when it announces its prices, but not in response to price cutting in the market. I think that Judge Posner's decision in Isaksen v. Vermont Castings, Inc. sets forth a reasoned approach to identifying an "agreement" in the post-Monsanto era. In Judge Posner's view, a showing of manufacturer efforts to persuade dealers to comply with minimum price levels, coupled with actual dealer compliance, is sufficient to establish an agreement to fix resale prices. Agreement may be communicated through conduct, as well as through words; an agreement can be "implicit, or signified by conduct in lieu of promissory language." Thus, a manufacturer's threat to a discounter to "mix up orders," followed by a price increase by the discounter because of the threat (albeit a year after the threat was made) was sufficient for a finding of an agreement. The basic point is that the Colgate doctrine permits suppliers to terminate discounters; it does not permit suppliers to persuade the discounters, through pressure or otherwise, to stop discounting. Once the discounters comply with that persuasion, we have a meeting of the minds or, in the parlance of Sharp, an agreement on price or price level. II. The Paradox of Resale Price Maintenance Now we will move on to the theoretical rationales for RPM. Let's start with the basics. Resale price maintenance is a vertical restriction by which a supplier controls the resale prices of its dealers. Minimum RPM, the more common form, requires dealers to charge prices at or above a certain floor level. Why would suppliers want to use RPM? It would appear that suppliers would ordinarily be interested in maintaining the lowest possible retail prices for their products, because low prices maximize their sales -- a low dealer margin is the equivalent of buying dealers' services cheaply. Moreover, a supplier should be able to exercise any market power it has by charging a high wholesale price without having to resort to RPM. Thus, supplier imposition of minimum RPM -- that is requiring dealers to charge premium prices -- seems peculiar at first glance as contrary to a supplier's self-interest. Many theories have been advanced purporting to explain supplier use of RPM -- some procompetitive and others anticompetitive. The theories that I will discuss today include several efficiency motivations, as well as supplier collusion, dealer collusion, and non-collusive theories of consumer injury. There is much disagreement in the antitrust community about the degree to which each of these theories underlies the actual use of RPM in the marketplace. This disagreement has resulted in a variety of RPM policy prescriptions, ranging from per se illegality to per se legality. While the actual motivations underlying individual uses of RPM are seldom obvious, the variety of circumstances in which RPM has been used suggests that each theory may be applicable under certain circumstances. It appears that no single theory can explain all, or perhaps even a majority, of the cases of RPM. Therefore, rather than attempting to determine the "real" motivation behind all instances of RPM, it may be more appropriate to acknowledge the possibility of several potential motivations. As the F.T.C. Bureau of Economics Report authored by Tom Overstreet notes, a per se standard could be the appropriate rule even if there are a substantial number of instances in which RPM is motivated by efficiency. If the efficiency-enhancing instances are relatively few, or if there is not a great efficiency loss if alternative mechanisms are utilized to accomplish the efficiencies, then net social benefits may accrue from a per se rule. Further, per se rules result in lower costs of litigation than rule of reason analysis, and per se rules can increase the predictability for businesses and enforcement agencies. It is beyond the scope of this speech to attempt to weigh into the controversy over whether RPM should be considered a per se offense, but it may make sense to step back and take a look at the various rationales that motivate a supplier to undertake the practice. III. Efficiencies While it is the anticompetitive use of RPM that underlies the current per se rule, I want first to discuss the efficiency theories, which have received the lion's share of the attention in recent years. Most commentators agree that RPM, at least in certain circumstances, can result in some efficiencies. The debate is over the degree to which efficiencies result, the instances where efficiencies appear more likely than anticompetitive results, the extent to which alternative means can achieve these efficiencies, and the appropriate legal rule given the possibilities of both efficiencies and anticompetitive effects. Proponents of RPM argue that the restraint is used to induce dealers to provide services. The RPM-imposed margin, they argue, is not a "profit" margin, but rather a payment from suppliers to dealers that enables dealers to cover their costs of providing sales and support services. For example, detailed product demonstrations by well-informed salespeople may in some instances be encouraged through RPM, and the provision of such services may be in the interest of both the supplier and the consumer. But why not simply contract with the dealers for such services? RPM proponents claim that in many instances such contracts may be costly to create, monitor, and enforce, and that service competition among dealers will generate the appropriate services more efficiently than explicit contracting. Difficulties in drafting and enforcing complex contracts requiring dealers to provide what often can be nebulous services (such as "attentive salespeople" or "attractive shopping environment") may strike a familiar cord with many of you. On the other hand, it is not intuitively obvious how providing a minimum dealer margin gives a dealer clairvoyance to divine the desired level of services. Perhaps, in some instances, dealers will have better knowledge of the necessary details of effective promotion than suppliers. But if service competition is the goal, why not simply allow dealers to compete with each other, and charge the margin that the market will bear for these added services? Economists, beginning with Lester Telser, claim that without RPM, dealers that do not provide these services may "free ride" on those that do. For example, a customer may go to a full-service dealer, take advantage of detailed product demonstrations, receive well- informed answers to technical questions, and be taught how to use the product. The customer then is free to go across town (or to a catalogue) to purchase from a discounter. Without vertical restrictions such as RPM, the theory goes, the full-service dealer cannot protect himself from such free riding, and thus may not have the incentive to continue to provide these costly services. Suppliers who consider such services important in order to raise the value of their products argue that it may not be possible to provide what they believe to be the best combination of price and service for their product without RPM. This free-rider theory may not have broad applicability, as many have noted, because outside of technical goods such as computers and stereo equipment, pre-sale services can be relatively unimportant. Nonetheless, RPM has been used in a wide variety of products including many in which costly pre-sale services appear less critical, such as apparel and drug store products. The use of RPM in such sectors may suggest anticompetitive motivations, although the economists tell us that this is not necessarily the case. Efficiency theories with broader applicability than the Telser theory have been offered. For example, Marvel and McCafferty have argued that "nontangible" services are offered by many upscale retailers, and that discounters could free ride on these services absent RPM. As they put it, many customers rely on upscale retailers to "certify" high quality and or high fashion. That is, the customer infers from the carrying of a certain line by an upscale retailer that the product has been pre-screened and concluded to be desirable from a quality or fashion standpoint. Such certification is arguably costly to retailers because they must hire sophisticated buyers and sample a broad range of potential sources in order to provide the latest in fashion to their customers. Absent RPM, customers could browse through an upscale store to get acquainted with the latest fashions, then make their purchases at discount stores. Although each case has to be analyzed in terms of its facts, it would appear, as a general matter, that big-ticket items with extensive pre-sale services such as consumer electronics are more susceptible to consumer free-riding than high fashion goods. Professors Scherer and Ross also express skepticism about the quality-certification theory. They point out that if such free- riding is limited in scope, high-amenity retailers will be able to charge prices sufficient to cover the costs of these amenities without the necessity of resale price maintenance. There exist, however, still other theories of efficiency- motivated RPM. One such theory argues that RPM creates incentives for dealers to provide services, but not by eliminating the ability of discounting dealers to free-ride on full-service dealers. According to this theory, dealers may not have adequate incentive to engage in activities, such as after-sale servicing, that the supplier considers important. Because the interest of dealers and suppliers do not coincide, the competitive process among dealers, absent RPM, may not result in certain important services being provided. Consider the incentive to provide after-sale service. Dealers that have not made long-term investments in establishing the quality and reputation of a particular supplier's product may not have a strong incentive to protect the goodwill of the supplier through after-sale service. This "bias" in the dealer's incentive is fueled by the fact that much of the adverse customer reaction to inadequate servicing may fall on the supplier. The distinction between this theory articulated by Klein and Murphy and the Telser theory is somewhat subtle. One way to think of the distinction is that the Telser story is a "horizontal" free-riding story, because RPM prevents certain dealers from free-riding on other dealers, while the Klein and Murphy theory is a "vertical" free-riding story, because RPM prevents dealers from free-riding on manufacturers. For example, if a product breaks down and the dealer does not adequately repair it, the theory goes, many customers will blame the manufacturer for producing a shoddy product, not the dealer for having poor servicing capabilities. If dealers are provided with an RPM-imposed margin, and are subject to being terminated if they do not provide adequate servicing, they may be induced to provide a greater supply of such costly services. However, dealers that provide inadequate service clearly will hurt their own business to some extent, and this incentive may be expected to be sufficient to induce adequate servicing in many cases. The vertical free riding theory may not be limited to just after-sale service, but may also be relevant for inducing promotional activities, such as providing additional sales effort or preferred shelf space. Given an RPM-imposed margin, dealers may be more willing to incur the costs of promoting a supplier's goods. For example, this theory has been said on occasion to underlie the use of RPM in the sale to wholesale distributors of certain perishable products, where refrigeration and product rotation can lead to spoilage and a loss of goodwill for the supplier. Dealers could "free ride" on the supplier by not handling it properly, and thus save the expense of such handling. RPM could provide dealers with compensation for the handling costs, and the threat of termination would further motivate the dealers to handle the product as specified by the supplier. One difficulty that I have in completely accepting this theory is that it conceivably could be applied to a broad range of industries, indeed any market in which dealer sales, promotion, pre-sale or after-sale servicing activities are valuable. Thus it appears to be a theory that would be easy for defendants to raise in instances where the true motivation for RPM is anticompetitive. This theory assumes that margins plus the threat of termination are necessary to induce adequate service. Query why in many instances the threat of termination alone doesn't suffice to achieve the desired level of service. Where it does suffice, imposition of RPM would appear less likely to have an efficiency rationale. The difference between this "vertical" free-riding theory and the quality certification theory illustrates how RPM may have efficiency justifications in industries with very different characteristics. Quality certification implies a dealer signal of quality which is valuable to manufacturers, while the vertical free-riding story implies a manufacturer signal of quality which is valuable to dealers. For some products, such as component stereo systems, manufacturer reputation serves as a strong signal of quality, but for others, such as fashionable dresses, dealers serve a relatively strong role in certifying quality. RPM, at least in theory, could increase efficiency in both types of markets. In addition to the services potentially associated with RPM discussed thus far, one dealer service of interest to manufacturers is carrying a large inventory of a supplier's goods. Dealers may not be willing to carry a large inventory of certain goods, such as new styles, when demand is highly uncertain. Suppliers may induce dealers to carry a large inventory -- with accompanying high costs of storage space, handling, financing, and risk of low demand -- by "guaranteeing" a certain margin through RPM. The costs of demand uncertainty is thus shifted from the dealer to the supplier, who may be more willing to take on such risks. In this way, product introductions, including procompetitive new entry, conceivably could be facilitated by this practice. While it is an interesting argument, I must admit I am skeptical about accepting it as a general proposition. Alternative means of compensating dealers for the risk of stocking newly introduced products that have less anticompetitive potential, such as a liberal returns policy, must be available in some instances. Another efficiency story has been advanced more by businessmen interested in instituting RPM than academics, but whether this increases or decreases the theory's credibility you can decide for yourself. The theory is that suppliers do not want their goods used as loss leaders, and prevent this by using RPM. But loss-leading would seem to increase the supplier's sales at no extra cost to the supplier, so one might think that loss leading is in the supplier's interest. The response is that such gains to the supplier are transitory; the practice ultimately leads many dealers that compete with the loss leading supplier to discontinue carrying the supplier's product, because they cannot sell it at a high enough price to make it worth the selling effort and shelf-space it requires. Reduced retail availability for the supplier's product may result, along with an accompanying loss of sales and reputation. Further, loss leading retailers may be low-quality retailers that are in effect "free riding" on the supplier's reputation, and this may result in the supplier's reputation being eroded because of its association with the low-quality retailer. I find this theory somewhat unconvincing because it requires that dealers be willing to discontinue carrying the very high- profile products that are attractive to use as loss leaders. Would a grocery be willing to discontinue carrying Oreo cookies or Kellogg's Corn Flakes even if sales of these products became unprofitable in a narrow sense? Surely dropping such high- profile products would adversely affect the grocery's sales more broadly. A similar sounding theory is that suppliers impose RPM as a way to encourage less efficient dealers to carry products in order to increase the retail availability and exposure of the product. While suppliers presumably have an interest in providing their products through the most efficient distribution channels, there may be additional channels that are less efficient, yet provide additional desirable exposure and availability to products. For example, a toothpaste manufacturer may most efficiently distribute its product through mass merchandisers and large chain drug stores. But making the product available at less efficient (at least with respect to the distribution of that particular product) corner drug stores and convenience stores may add substantially to the exposure of the product and thus its sales. Without RPM, according to this theory, these outlets may be unwilling to stock the toothpaste, because the higher prices they must charge compared to other retailers will lead to limited sales. By leveling the playing field for the different categories of retail competitors, RPM can provide a stronger incentive for these less efficient outlets to carry the product. Thus in effect, RPM results in a marketing service for the manufacturer, which may increase his ability to compete with other toothpaste manufacturers. These last two stories seem necessarily to imply higher prices for consumers without any apparent additional customer service, unlike the other theories, which are based on the idea that the manufacturer is buying services for its customers from its dealers. But the theories also imply wider availability of the product. The wide availability and exposure of the product (which could be services valued by consumers) are a means of marketing the product, and perhaps such marketing efforts can be procompetitive. Yet the wider availability comes at a cost. More efficient dealers are prevented from passing on cost savings to the customer. One problem with many of these efficiency stories is that RPM, by itself, does not provide a direct incentive for dealers to provide the particular services that suppliers want, whereas explicit contracts with dealers would. For example, a supplier may wish for its dealers to demonstrate products, but with RPM, dealer competition may result in dealers "spending" their RPM- induced margin on discounted accessories instead. Only if nonprice competition can occur solely through the means that suppliers want to encourage would RPM alone necessarily result in the desired services. Thus in addition to RPM, suppliers must continue to monitor dealers' provision of services, and perhaps threaten dealer termination if services aren't adequately provided. RPM's ability to reduce costly contractual problems between suppliers and dealers appears less impressive when viewed in this context. While many of the efficiency stories have intuitive credibility, they are generally not well-documented and efforts to apply them to some industry contexts seem strained. For that reason, many remain skeptical of the prevalence of RPM-induced efficiencies. The story we hear most often from RPM proponents in response to expressions of skepticism about these efficiency stories is that if RPM is not used as a part of a horizontal collusive agreement, it must be efficiency motivated, or else it would not be in the interest of a supplier to impose the scheme unilaterally. Why would suppliers impose a margin on their distributors if not to obtain some service benefits? Higher retail prices will reduce sales without generating profit for the supplier. Thus, if the supplier can exercise unilateral market power, it will be preferable to raise the wholesale price rather than in effect to "give" the monopoly profits to dealers through an RPM scheme. To my mind, this argument still leaves unanswered the question of just what services are being provided for, say, a pair of Jordache blue jeans at Saks Fifth Avenue that aren't also provided at K-Mart? As Bob Pitofsky has asked, don't blue jeans in both stores get laid on a table, picked up by customers and taken to the dressing room, tried on and purchased? I will venture an answer, admittedly with some skepticism in my own mind, at least to its general applicability. You can judge for yourself its degree of plausibility. Perhaps at Saks the salespeople are more numerous and courteous, as well as more knowledgeable about fashion. Perhaps Saks keeps a broader inventory of each style, size and color in the Jordache line, so that they almost always will have the Jordache item that a customer wants. Perhaps the lines at the cash register are shorter at Saks, because they are heavily staffed so that customers will not have to wait as long. Perhaps the dressing rooms are more pleasant than those at K-Mart. Perhaps even the table on which the jeans are laid at Saks is more expensive than the table at K-Mart. And perhaps Saks keeps only the latest fashions on its shelves, so that customers obtain some assurance of high fashion by shopping there, even if they haven't studied the latest fashion magazines. Even so, however, the question remains as to how RPM in fact achieves the desired level of service. The fact that both high and low end retailers sell many types of products at near uniform price suggests that setting the resale price is not of talismanic significance in causing all retailers to provide a uniform level of services. IV. Anticompetitive Theories Most commentators also agree that resale price maintenance can be used to facilitate collusion, either among suppliers or among dealers. Indeed, it is because of the collusive use of RPM that the Supreme Court, in its Sharp decision, rationalized continued per se treatment of RPM, and distinguished non-price vertical restraints which are viewed under the rule of reason. Let us look at supplier collusion first. How can RPM facilitate collusion? For a collusive agreement among suppliers to be successful, it must be possible for the conspirators to detect cheating. When suppliers sell to dealers who then resell the goods, it often may be easier to monitor the retail price than the wholesale price. By fixing a minimum retail price through RPM, a supplier's incentive to cheat on the agreement by cutting the wholesale price is reduced, because the dealer, not the supplier, will reap the profits from the price cutting. Simply stated, RPM can aid in the detection of supplier cartel cheating, as well as reducing the incentive to cheat, and thus can facilitate the maintenance of collusive pricing. But it should also be noted that RPM does not necessarily offer a panacea that enables suppliers to collude in spite of other industry characteristics that inhibit collusion. For example, if dealers can affect a supplier's sales substantially through their actions -- for example by giving the supplier's product preferred shelf space or adding value to the product -- a supplier could still profit by cutting price to a dealer, and ultimately such actions may undermine the collusive agreement. In many cases, however, these actions of the dealer will also be detectable by the supplier's competitors, and thus continue to signal to them that a rival may be cheating on the cartel agreement. Now let us turn to the possibility of RPM facilitating a collusive agreement among dealers, either at the retail or wholesale level. Dealer competition often is characterized as fiercely competitive, and thus a facilitating device such as RPM may be necessary in attempts to collude. If dealers cannot police an agreement among themselves, they may induce suppliers to impose RPM, and thus to act as the cartel "enforcer." One variant of dealer collusion we would look for entails dealers, facing the prospect of a new form of competing and more efficient distribution, inducing suppliers to impose a minimum price through RPM. Any entrant would then be prevented from being able to pass on its efficiencies and the incumbents would be insulated from competition that ultimately would benefit consumers. For example, it has been noted that the rise of mass merchandisers, supermarkets, and department stores seems to be responsible for a rise of support for RPM from more traditional retailers. Empirical studies suggest that some dealers in the past have joined together, often through trade groups, to pressure manufacturers to institute RPM. Thus, although dealer cartels may appear at first blush to be uncommon, because retail and other distribution markets tend to be more atomistic and entry generally is easier into retail markets, the empirical evidence suggests that dealer cartels are a valid concern. As noted at the outset, it would not appear to be in the interest of suppliers to have their dealers collude, because higher dealer profit margins imply lower sales and lower profits for the supplier. Thus, it appears that dealers must have a credible economic threat over suppliers, such as the threat of a boycott. If suppliers have alternative distribution channels available to them, dealers will have little economic leverage over suppliers. Further, an RPM scheme imposed by a dealer cartel may not work if some suppliers that sell competing products do not participate in the agreement. Thus with this collusion theory, "cheating" at both the supplier and dealer level is possible. These theories of supplier and dealer cartels may appear to beg the question, "Why is a legal rule against RPM necessary when the underlying collusion already is illegal?" Let me offer two answers. First, the collusive pressures leading to the RPM in many instances may not be detectable or actionable, but RPM frequently is both. As Professor Areeda has argued, informal pressures and "quasi-collective" actions can bring about anticompetitive RPM in instances in which a strict conspiracy standard may not be met. Parallel action by powerful retailers that falls short of a legally-cognizable conspiracy may produce the same results as a horizontal cartel -- the unjustified elevation of retail prices obtained through pressure on the supplier. Second, non-collusive theories of consumer injury, to be discussed next, may have some validity. Theories of how RPM can be harmful to consumers when not used as a means of facilitating a dealer or supplier cartel have not been as widely discussed or accepted as the collusive theories. But these theories appear worthy of consideration. William Comanor has argued that while RPM may well induce the provision of services, many customers may be stuck with paying a higher price for services they don't want. Even if these customers do not consume these services, they are "forced" to pay for them if an unbundled package is not available. For example, a large number of computer aficionados may need little or no pre-sale service because of their expertise. Computer novices, on the other hand, may place high value on these services. If personal computers are not available without RPM- induced services, the aficionados (the so-called inframarginal consumers) must pay for services they don't want. The economic cost to the aficionados may more than offset the benefits provided to these other customers who value these services. But while RPM conceivably can have a welfare-reducing effect in this way, Comanor's argument has come under some criticism. First, if a large number of consumers do not demand the RPM- induced services, and the supplier faces competitors, those competitors can be expected to provide low price/low service alternatives. Second, this theory begins with the presumption that the services induced by RPM are valuable to some consumers, thus leaving us with the difficult task of judging in which cases the services are excessive. Further, Comanor's theory is potentially not limited to RPM, but is applicable to a broad range of business decisions about product features and quality. For example, a supplier of luxury automobiles may include as a standard feature in all its cars a powerful stereo that is not valued highly by some of its customers. The provision of such "excessive quality" is difficult to distinguish from RPM-induced excessive services. Comanor's theory is therefore intriguing, but I am not entirely sure how much practical antitrust application it has. Bob Pitofsky has argued that the intra-brand competition prevented by RPM may be more critical to consumers than any services encouraged by RPM, at least in some cases. Suppose that K-Mart and Saks both carry Jordache jeans; K-Mart's pricing should apply some competitive pressure on Saks. RPM would reduce the degree of competitive pressure that K-Mart could exert on Saks. Furthermore, Pitofsky argues that this intra-brand pricing pressure at the retail level can place pressure on Jordache to lower its wholesale price. It is difficult for many skeptics of the RPM efficiency stories to believe that such intra-brand competition is not more valuable than the services that might be encouraged through RPM. Indeed, Pitofsky also points out that an additional competitive benefit that may be suppressed by RPM is the ability to pass on to consumers any efficiencies realized by more efficient dealers. But is intra-brand competition needed to prevent anticompetitive pricing by Jordache? Wouldn't inter-brand competition, from such sources as Calvin Klein and Gloria Vanderbilt, if not also Levis, Wrangler, and Lee's, exert substantial pressure on Jordache's prices? Even so, there could still be a price premium that might be eliminated by intra-brand competition. In highly differentiated and highly advertised product markets, it is certainly possible that suppliers are distant enough substitutes from each other that some premium can be charged. But then if this is the case, why would a supplier with market power in effect give some of its profits to dealers through RPM? Perhaps the dealer has some market power itself with respect to the supplier, and the two negotiate a "bilateral monopoly" agreement whereby they share the monopoly profits through RPM. Professor Areeda discusses how powerful dealers may individually be able to induce RPM from suppliers. But why would a dealer with power over its supplier choose RPM rather than simply negotiating lower wholesale prices with the supplier? Note that a dealer would not want RPM to be imposed just on itself without the restraint also being imposed on its rivals, or else it would be at a pricing disadvantage. The dealer must have sufficient power to induce the supplier to impose RPM throughout its market. It might appear that a powerful dealer could induce a larger margin if it negotiated for a discounted wholesale price only for itself, rather than requiring that an RPM-imposed margin be placed on its competitors as well as itself. Areeda answers that a dealer may realize that some of its competitors may have a similar ability to obtain lower wholesale prices. For that reason, competition in the retail market might prevent the dealer from realizing profits from the lower wholesale price. With RPM, the dealer margin is protected. Dealers therefore might prefer RPM to lower wholesale prices. Note how this theory, though it does not require explicit dealer collusion, is driven by the actions of dealers who have some degree of market power. Some amount of dealer power thus appears to be a common denominator for many of the anticompetitive theories of RPM. Lastly, perhaps a supplier may in some instances guarantee an RPM-imposed profit margin to a dealer in exchange for the dealer not carrying competitors' products. Perhaps the dealer's services are crucial for suppliers in the market. RPM could be the quid pro quo for an exclusive dealer agreement that makes entry into competition with the supplier more difficult. In conclusion, it appears that consumer injury may occur from both collusive and non-collusive uses of resale price maintenance. V. Maximum RPM Before closing, I want to raise one additional issue regarding RPM for which the proper policy prescription is much less controversial. The Supreme Court's recent decision in Atlantic Richfield Co. v. USA Petroleum Co. suggests that the day may not be far off when we witness the demise of the per se prohibition against maximum RPM established in Albrecht v. Herald Co. Moreover, the proposed legislation concerning RPM that would, among other things, codify the per se rule against RPM by its terms, would not apply to maximum RPM. These are not trivial developments given that approximately 30% of private RPM litigation in a recent sample were maximum RPM cases. The Court directly observed in the Arco case that maximum RPM may be used to counteract any market power that a distributor may gain from exclusive territorial agreements. As the Court noted, in those circumstances, maximum RPM tends to "drive prices toward the level that would be set by intense competition." Maximum RPM may also be useful to manufacturers for ensuring that their promotional discounts are passed on to consumers. One point, however, bears further mention. Maximum RPM may be anticompetitive when it is used as a subterfuge for minimum RPM. It is the economic effect of a pricing policy that should control, not the label chosen by the one imposing the restraint. But this caveat should not undercut the basic thrust of Arco. And it appears that Arco signals a willingness to analyze maximum RPM under the rule of reason. Perhaps the most significant hint of the Court's readiness to do so was its statement that the "procompetitive potential of vertical maximum price restraint is more evident now than when Albrecht was decided, because exclusive territorial arrangements and other nonprice restrictions were unlawful per se in 1968." This statement seems to demand reconsideration of the per se prohibition in light of the new learning. Moreover, this is an issue on which there seems to be wide agreement embracing both proponents and opponents of the other flavor of RPM -- minimum price restraints. Lastly, an insight of Judge Posner in a maximum RPM case that predated -- and perhaps also presaged -- Arco, seems worthy of note. In this case, a manufacturer, who had given its deal- ers exclusive territories, ran special promotions several times a year, when it offered and advertised special deals on its product. During these promotions, it would give its dealers a price break. To make sure that the dealers passed on the price break to consumers, it threatened to terminate dealers who sold the product at a price higher than the advertised price. According to Judge Posner, If it is lawful to advertise a retail price, it should be lawful to take at least the minimum steps necessary to make that advertising benefi- cial. It would be pretty embarrassing for a manu- facturer who had advertised a special retail price to be bombarded by complaints from consumers that dealers were refusing to sell to them at that price. Such refusals would make the advertising misleading and might even expose the manufacturer to sanctions under the Federal Trade Commission Act or counterpart state regulations. So if re- tail price advertising by the manufacturer is to be feasible the manufacturer must be allowed to take reasonable measures to make sure the adver- tised price is not exceeded. Judge Posner offers a highly persuasive and eloquent explanation of the procompetitive uses of maximum RPM. The short of the matter is that maximum RPM, to go back to the title of this speech, is much more likely to be consumers' friend than their foe. It offers a mechanism for ensuring that manufacturer discounts are passed on to consumers. Thus the mounting indications that maximum RPM should not be automatically condemned portend a positive development for consumers. VI. Final Comments In these remarks, I have outlined a variety of theories concerning resale price maintenance, both procompetitive and anticompetitive. There seems to be agreement that both efficiencies and anticompetitive effects result from RPM in some instances. But proponents of efficiency theories argue that RPM is predominantly beneficial to consumers and proponents of the anticompetitive theories argue that RPM is predominantly harmful. Not sharing the confidence of either camp, I will not venture a similar leap of faith today. Thank you for the opportunity to discuss resale price maintenance with you. I will be happy to take some questions.