I. Introduction
My topic today is merger enforcement, and particularly merger litigation, at the FTC. As suggested by three preliminary injunction cases that were litigated simultaneously over the summer, FTC v. Cardinal Health, Inc., FTC v. McKesson Corp., and FTC v. Tenet Healthcare Corp., the Commission is litigating more cases these days, and investigations overall seem to be more complex than ever. This morning I want to review some of these cases and discuss some of the important aspects of the courts' decisions. I will then elaborate on three areas of merger enforcement: coordinated interaction, mergers in high technology markets, and vertical mergers, and use some FTC enforcement actions to illustrate our analysis. As always, my remarks today are my own and do not necessarily reflect the views of the Commission or any of its Commissioners.
Let me begin with a review of some important statistics. The merger wave, which has been the subject of recent Congressional hearings, has had a substantial impact on the work of the Bureau. Since the beginning of fiscal year 1998 we have received over 4400 merger filings, approximately 28% greater than fiscal year 1997, and the largest volume of merger filings in history. It is certain that 1998 will be the seventh consecutive yearly increase since 1991, when there were 1,451 filings, a time when our staffing was almost identical to what it is today. In other words, we are reviewing three times as many filings with the same staffing we had seven years ago.
To date in fiscal year 1998, the Commission has initiated 23 merger law enforcement actions, including three preliminary injunctions and 20 consent agreements. An additional six transactions were abandoned in the face of probable enforcement action. Moreover, we are using our scarce resources well by targeting second requests to those acquisitions in which we ultimately find enforcement action is necessary. In the past year, we issued 38 second requests, which is actually fewer than in earlier years, when there was a smaller number of overall transactions.
II. Merger Litigation
The most notable aspect of antitrust merger enforcement recently is the increasing amount of merger litigation. At the Commission, we recently successfully litigated two merger cases simultaneously: our challenge of Tenet's acquisition of a hospital in Poplar Bluff, Missouri and the consolidated challenge to two mergers involving the four largest drug wholesalers in the United States. These cases were substantial, not only in terms of the length of the trials, but also the resources devoted to litigating them. To date in Fiscal Year 1998, the Bureau of Competition has spent more hours on merger litigation than in any other year in recent memory, over 40 person years, and approximately more than four times as many hours as in earlier years.
Let me pose a question -- Why is there an increasing level of litigation? Have the enforcement agencies raised the standards? Are they reinterpreting the law or pushing the envelope in an effort to develop new law?
I do not think that anything that dramatic is occurring. We continue to enforce the law as written and as interpreted by the courts. The Merger Guidelines set forth our method of analysis. We have no preconceptions of how any particular market operates. Rather, with each new case, we must roll up our sleeves and search for the facts that will guide our understanding of the market.
Two trends may be contributing to the increased level of merger litigation. First, as others have observed in greater detail,(1) unlike the transactions of a decade ago which were primarily motivated by financial considerations, today's transactions are increasingly strategic in nature. Many firms perceive a need to be the market leader, and use acquisitions as the path to leadership. Other acquisitions come about where firms seek to acquire critical inputs in the market. Other acquisitions may be motivated by a desire to relieve competitive pressures in the market.
The three cases we recently litigated are representative of these trends. While many hospital mergers are efficiency-driven and procompetitive, a small number, such as the one in Tenet Healthcare, threaten to increase the firms' market power and ability to raise prices. Hospital merger cases can be difficult to litigate, especially on geographic market issues, and Tenet Healthcare was no exception. The drug wholesale cases similarly illustrate the complexities of the contemporary business environment in healthcare. The drug wholesaling industry had experienced considerable consolidation over several years, resulting in lower costs and greater services for customers. At some point, however, the market power concerns resulting from consolidation begin to outweigh the claimed efficiency benefits.
Second, I think the enforcement agencies increasingly are recognizing that not all cases can be settled, and that a settlement which does not effectively restore competition is not worth adopting. Simply, there are some transactions that can not or should not be resolved through some type of remedy short of a total injunction. Our obligation is to ensure that competition is restored to the level that existed prior to the merger. Frequently, relief that does not include divestiture of a facility is inadequate. For example, a licensing arrangement may not fully enable a new firm to become an aggressive rival, because it remains in a dependent relationship with the merged firm. In the past few years the Bureau of Competition has systematically reviewed the results of our divestitures and we have learned that they sometimes do not adequately replace the competition eliminated by the transaction.
On occasion, parties to a transaction propose that a merger be resolved through the sale of some package of assets, which have the appearance of being pieced together to create a new rival in the market, or expanding a fringe player. Often these proposed divestitures are rejected, because an important rival will be eliminated by the transaction and the divestiture of some set of assets will not fully restore competition. Our obligation is to restore the level of competition that existed prior to the merger, and that will require a sufficient set of assets to create a firm that is the competitive equal of the firm eliminated from the market.
The drug wholesale cases provide an illustration. As anyone who followed the trial knows, the court explored every opportunity with the parties to find a settlement that could permit the proposed mergers to go forward. The parties suggested that a divestiture of several drug wholesale distribution centers would be sufficient to restore competition. As we told Judge Sporkin, such a divestiture would have been severely inadequate, because in our view, customers increasingly demanded firms that could provide national service and divestiture of a handful of distribution centers could not compensate for the loss of these national competitors that would have resulted from the proposed transactions. Divestiture would have eliminated two firms -- Bergen and AmeriSource -- that had provided intense competition and innovation, and that would be lost forever.
We recognize that merger litigation may well be an increasing part of our enforcement efforts and began several years ago to improve our ability effectively to litigate cases. We have increased the level of training, especially in the litigation area. Much of our training efforts focuses on new or relatively junior attorneys, to facilitate their hands-on involvement in litigation. We believe that our successes inCardinal Health and Tenet indicate that our increased emphasis on trial advocacy training is paying off. Judge Sporkin's recognition of the "excellent" performance of "other capable lawyers from [our] offices" confirms to us, and should confirm to our potential adversaries, that the Commission stands ready to go to court when necessary to protect the public interest in competition -- and will do so, and will succeed.
McKesson/AmeriSource and Cardinal Health/Bergen Brunswig.
A particularly significant merger enforcement action was our challenge to two mergers involving the nation's four largest drug wholesalers -- McKesson merging with AmeriSource and Cardinal Health with Bergen-Brunswig. If the mergers had been permitted, the two survivors would have controlled over 80% of the prescription drug wholesaling market, significantly reducing competition on price and services. We filed the two actions in district court in March, and the case was litigated for approximately seven weeks during June and July. Judge Sporkin enjoined both acquisitions in a 73-page opinion issued at the end of July.
There are many substantive issues addressed by Judge Sporkin's opinion. Today, I would like to address five of them: relevant market analysis, entry barriers, power buyers, efficiencies, and regulatory relief.
Relevant market. As in most merger cases, definition of the relevant product market was a hotly contested issue. The FTC alleged that the relevant market consisted of the cluster of services provided by drug wholesalers to institutional customers (e.g., hospitals) and retail pharmacies, including warehousing, distribution and other value-added services. Drugs sold through wholesalers accounted for $54 billion in sales in 1997, out of a total $94 billion in industry sales (which included drugs sold directly by manufacturers to consumers). The defendants understandably argued for a broader market that would include other means of distribution, such as direct purchases from manufacturers and self-warehousing.
The court dealt with the issue in a way that is economically sound, firmly grounded in Supreme Court precedent, and well supported by the facts. The key question was whether customers had economically viable substitutes for wholesale distribution services? The court found that hospitals and independent drugstores would not turn to purported substitutes -- primarily direct delivery and self-warehousing -- to defeat an anticompetitive price increase. There was a good, common-sense reason for that, well-supported by the evidence. Hospitals have increased their reliance on wholesalers, at 80% of their needs in 1997, and that amount has been growing over the past ten years. That was not likely to change even if the cost of wholesale distribution services increased by a small amount. Even though the court found that large retail chains could increase their self-warehousing, they represented a small segment of retail chain pharmacies. The court found that self-warehousing by these chains was insufficient to protect independent pharmacies and hospitals which could not switch to other means of distribution and would be harmed by a reduction in wholesaler competition.
In sum, consistent with the teachings of Brown Shoe, and the Merger Guidelines, the court looked beyond mere functional interchangeability to practical economic interchangeability. A relevant market need not be airtight, or defined with absolute precision. The existence of some functional substitutability does not broaden the product market if they are not economically viable alternatives.
Entry barriers. Usually a court's analysis of entry is sort of a binary exercise. In most merger decisions, a court has either found that entry is timely, likely and sufficient, or it is neither timely, likely, or sufficient. Judge Sporkin's opinion is one of the few that finds that entry could satisfy some conditions but not others, but ultimately concludes that entry could not avert the anticompetitive effects of the mergers.
Judge Sporkin found that entry could be timely, observing that firms could acquire the necessary personnel and distribution facilities because the merger would result in the consolidation of numerous distribution centers. But he agreed that there had been little entry historically, and that the continual decline of profit margins suggested that new entry or expansion was unlikely. As to expansion of fringe firms, the court found the solution to be more "uncertain" than defendants suggested but nonetheless a real possibility.
The court's decisive finding on this issue was that entry or expansion would be insufficient to avert the anticompetitive effects of the mergers. Although small regional wholesalers had taken some business from the national firms, the court concluded that they were not a sufficient competitive deterrent and lacked the national presence necessary to effectively restore the lost competition.
Judge Sporkin's opinion, like the court's opinion in Staples, helped to reconcile the court's decision in Baker Hughes with the Merger Guidelines. Some have suggested that the decision in Baker Hughes enables a defendant to defeat the government's case simply by showing the lack of insurmountable entry barriers. The decisions in Staples and Cardinal make clear that the plaintiff need not demonstrate the existence of insurmountable entry barriers, but rather the inquiry focuses on an overall assessment of whether entry will be timely, likely, and sufficient.
Power buyers. Every few years a new merger defense seems to become popular and a few years ago it was the presence of power buyers. A number of courts, including the court in Baker Hughes, relied on the size and sophistication of buyers to reject the government's request that a merger be enjoined.
In these cases, the defendants relied heavily on the fact that some of the large chain drug stores were large buyers, had the ability to increase self warehousing and were sophisticated purchasers. So were the group purchasing organizations acting at the behest of hospitals and independent pharmacies. The court acknowledged that "the customers of the Defendants possess a significant amount of leverage in contract negotiations. On the whole, buyers monitor prices very closely and are aware of the wholesalers' individual cost-structures." But because there were many buyers including independent hospitals and independent pharmacies that could not exercise similar leverage, the buyer power argument was insufficient to alleviate the court's concerns.
Efficiencies. Judge Sporkin's decision may well be a model that future efficiency decisions are likely to resemble. Although the court recognized that there are decisions that have rejected consideration of efficiencies, it followed the analysis set out in the 1997 revisions to the Merger Guidelines.
Among the efficiencies claimed by the parties were (1) distributional efficiencies through the closing of overlapping centers; (2) superior purchasing practices; (3) buying power; and (4) reduction in overhead and inventory costs. However, Judge Sporkin found that the evidence strongly suggested that many of the efficiencies could be produced in the absence of the merger, and that the parties would pass only 50% of the savings on to consumers (instead of their historical average of 80%). Ultimately, he concluded that although there would be some efficiencies, they were insufficient to overcome the anticompetitive aspects of the transaction.
Perhaps the most important aspect of the efficiency analysis is the court's consideration of the role of excess capacity. Although elimination of excess capacity could have resulted in some cost savings, the critical question, as posed by the revised efficiency section of the Guidelines, is how the proposed efficiencies affect competition in the market. In its examination of competitive effects, the court recognized that competition, and therefore consumers, would not benefit from the reduction in excess capacity. Since excess capacity was the catalyst for aggressive competition, the court concluded that the "mergers would likely curb downward pricing pressures and adversely affect competition in the market." This recognition of the effect of the efficiency on post-merger competition is an important insight in merger analysis.
Regulatory relief. Finally, one issue that arose during trial was whether there was some form of regulatory relief that could be imposed to ameliorate the anticompetitive effects and to permit the mergers to occur. The parties pledged not to increase prices and to pass on 50% of any costs savings resulting from the merger. We argued that involving the court as a regulator of prices would have been a "second best" solution to continued competition among the four firms, would have been unsound antitrust policy, and was contrary to law.
The court ultimately rejected the defendants' promise as an antidote to anticompetitive effects because resorting to a "price cap" would have effectively deprived consumers of the lower prices that would derive from competition. In this market competition continually led to lower prices. The court relied on evidence relating to McKesson's first attempt to acquire AmeriSource in 1988. If that acquisition had been approved based on a "price cap" promise, consumers would have been deprived of substantial decreases in prices that occurred over the next decade.
Two final observations about the case. First, the court found that the drug wholesale market was very competitive, with firms working aggressively to grow and become more efficient. As the court noted, this market is an "excellent example of how effectively the nation's market system works." This level of competition made our enforcement action all the more appropriate. The benefits of competition are substantial; and we estimate that our enforcement action will save U.S. consumers over one hundred million dollars annually.
Second, in the past some have criticized the enforcement agencies' use of consents rather than litigation as diminishing the role of the courts in merger enforcement. One perceived problem of the use of consents is that there is relatively little development of case law on mergers. One benefit of litigating Staples and Cardinal is that the decisions have helped to clarify the law, and reconcile it with the Merger Guidelines on issues such as entry and relevant product market. This ultimately may play an important role not only for merger enforcement, but also for the development of antitrust law generally.(2)
Tenet/Poplar Bluff
In July, the FTC and the State of Missouri obtained a preliminary injunction temporarily preventing the proposed merger of the only two commercial hospitals in Poplar Bluff, Missouri. Tenet Healthcare Corporation, the second-largest hospital chain in the country, owns Lucy Lee Hospital, a for-profit general acute care hospital in Poplar Bluff. Tenet proposed to acquire Doctors' Regional Medical Center, a physician-owned for-profit hospital and, apart from a hospital run by the Veteran's Administration, the only other general acute care hospital in Poplar Bluff.
This was a far shorter trial than the drug wholesale case, accounting for only five days of hearings. The litigation primarily focused on three issues: geographic market; whether DRMC, while not failing, was a "flailing" firm; and efficiencies. The court found for the FTC on each of these issues. Tenet and DRMC have appealed the order granting the preliminary injunction to the U.S. Court of Appeals for the Eight Circuit. The FTC has initiated a Part III proceeding seeking to enjoin permanently the proposed acquisition, which has been stayed pending resolution of the appeal to the Eighth Circuit.
Let me address two aspects of the decision that I think are important outside the hospital merger context: the "flailing firm" defense and efficiencies.
Flailing firm. Judge Perry's opinion is fully consistent with the method of analysis outlined in the Merger Guidelines. The Guidelines, like the substantial majority of the small number of cases on point, do not recognize a "flailing firm" defense as such. The standard for the failing firm defense set forth in the Guidelines requires the firm to demonstrate that it would not be able to reorganize successfully under the Bankruptcy Act. In other words, the analysis is not static, considering only whether the firm is experiencing current difficulties. Rather, the analysis is dynamic, asking whether the firm is capable of changing and adapting so as to be able to compete successfully in the future.
While admitting that it did not satisfy the standards of the failing firm defense, DRMC claimed that excess capacity, declining admission and occupancy rates and deferred capital improvements spending threatened the hospital's long-term viability. Judge Perry declined to accept at face value DRMC's claim that it was doing less well than it had in the past, and would continue to decline in the future. A declining inpatient occupancy rate was not dispositive since increased flexibility in the provision of inpatient services and the growth in outpatient services have rendered traditional measures such as occupancy rates much less meaningful than they were in the past. Likewise, efforts to control costs, such as changes in Medicare/Medicaid reimbursement rates and increases in managed care programs would not necessarily cause DRMC to lose its relative share of the market. Ultimately, neither factor was sufficient to suggest that DRMC would not continue to be a competitive force in the market.
Efficiencies. The defendants suggested that the merger would allow them to combine two inefficient, underutilized hospitals. The court rejected these claims for three reasons. First, the hospitals were being operated on an efficient basis. Second, if there was excess capacity, the hospitals could rationalize capacity and achieve savings even without the merger. Finally, even if the hospitals were able to achieve efficiencies, they would not likely benefit consumers because the merged hospital was unlikely to pass any cost savings on to consumers absent competitive pressure to lower prices. As the Guidelines observe: "Efficiencies almost never justify a merger to monopoly or near-monopoly."
In short, Judge Perry's opinion confirms that a failing firm defense and an efficiencies claim must apply a dynamic rather than a static analysis. It is not enough to say that recent or anticipated changes in the market have caused or will cause a firm's absolute performance to decline, or that a firm's performance after the merger will be better than its current performance. The relevant questions are whether and to what extent the firm can independently adjust to market changes, and whether market changes will affect the firm's competitive position relative to its competitors. Likewise, the argument that a transaction will result in efficiencies is not persuasive absent an analysis of efficiencies that could be obtained absent the transaction and whether the efficiencies in question will improve the competitive performance of the market.
III. Coordinated Interaction Cases
For those who may think that we only challenge horizontal mergers under a unilateral effects theory, the decision in the drug wholesale cases proves otherwise. I will discuss the coordinated effects analysis in that case and in three recent merger consents: first, the joint venture between Shell and Texaco;(3) second, Degussa Aktiengesellschaft's proposed acquisition of E.I. DuPont's worldwide hydrogen peroxide business;(4) and finally, the proposed joint venture between Exxon and the Royal Dutch Shell group of companies to develop, manufacture and sell fuel and lubricant additives.(5)
Drug wholesalers. In drug wholesalers, one of the factors the court relied upon in assessing competitive effects was the past efforts of coordination. In particular, the court focused on contracts between VHA, one of the largest hospital GPOs with three of the four defendants -- Cardinal, Bergen, and AmeriSource, which contained certain provisions that on their face guaranteed VHA members a favorable price. There was evidence, however, that in practice, the clause set a floor on the price that the firms would offer to other hospitals and GPOs and the agreement served to police the defendants from trying to undercut one another. Based on this evidence, the court concluded that the defendants would have an increased ability to coordinate their pricing practices after the merger.(6)
Shell-Texaco. The Shell-Texaco joint venture, which combined those firms' refining and marketing assets, raised several interesting issues in diverse markets. The Commission entered a consent order that required the divestiture of an interest in a petroleum products pipeline, a Washington state refinery, terminals and retail assets in Hawaii, and retail assets in San Diego. The primary focus of our competitive concerns in refining, terminaling, and marketing was coordinated interaction. In fact, in some of the markets the HHIs were less than 2000. In one market, the refining of gasoline for California, the transaction would have raised the HHI by 154 points to 1635, within the moderately concentrated range.
Although the concentration numbers in gasoline refining may not have been as substantial as in other merger enforcement actions, the evidence suggested a significant threat of coordinated interaction. In gasoline refining, the products are homogeneous, and wholesale prices are publicly available and widely reported to the industry. Refiners therefore readily can identify firms that deviate from a coordinated or collusive price. Existing exchange agreements between refiners likely would facilitate identifying and punishing those deviating from a coordinated or collusive price.
One critical fact was that industry members have raised prices in the past by selling products outside the market, sometimes at a loss, in order to remove supplies that had been exerting downward pressure on prices. That type of conduct would not make economic sense from the perspective of an individual firm unless it could be confident that it would lead to a coordinated increase in price.
Degussa. The facts in Degussa were similar in some respects, but the market was somewhat more concentrated and big purchasers were involved. Hydrogen peroxide is an inorganic chemical that is used in a variety of applications as an oxidizing agent to encourage different chemical reactions. The paper and pulp industry is a major user, as are the textile and mining industries. The proposed transaction between Degussa and E.I. DuPont would have reduced the number of North American producers from seven to six, and increased the HHI by over 500 points, to over 2500. The product market was well-defined -- there are no cost-effective substitutes for hydrogen peroxide for these applications -- and the geographic market was confined to North America due to transportation costs and other constraints.
In addition to structural conditions (i.e., high concentration and restrictive entry conditions) and inelastic demand, there were a number of other factors that made coordinated interaction -- i.e., collusion -- a serious concern:
First, hydrogen peroxide is a highly homogeneous product that is purchased primarily on the basis of price. That makes it easier to agree on price, and it also makes it easier to detect cheating. Second, reliable pricing information is available to the producers due to the use of delivered pricing, the practice of advance announcement of price increases, and customer arrangements including meet-or-release clauses. Thus, the firms have good information on what their competitors are doing, and what they're planning to do with respect to price.
Third, there was a past history of express collusion among hydrogen peroxide producers in Europe from the early 1960s through the late 1970s. The firms involved were producers that after the acquisition would be the leading manufacturers in North America.
Fourth, there are industry practices that can facilitate interdependence and coordination in a concentrated market. For example, producers make sales to each other with some frequency. At times those sales appeared to be intended to avoid competitive conflicts among producers.
Fifth, producers have been able to maintain large differentials in pricing among different end-uses, even though the product is essentially indistinguishable in its performance characteristics. That indicates at least two things: an ability to coordinate prices, and the inability of arbitrage to break down pricing differentials.
Finally, company documents projected higher prices as a result of the proposed acquisition. In general, that kind of evidence is consistent with both a unilateral effects and a collusion theory, but here the structural conditions indicated that collusion is the more likely outcome.
Thus, in several respects, Degussa is a classic example of an industry ripe for collusion. I don't want to suggest that all of the conditions found in that case, or even a majority of them, have to be present for a coordinated interaction case. That is illustrated by the proposed joint venture between Exxon and Royal Dutch Shell.
Exxon/Royal Dutch Shell. The Exxon/Royal Dutch Shell joint venture would have combined two existing businesses engaged in producing and selling a viscosity index improver ("VI improver"), which is used in the production of motor oil to enhance its overall viscosity and ability to protect moving parts. It represents only a small percentage of the cost of the end product, but it is an indispensable ingredient. The transaction would have decreased the number of merchant producers from four to three, and there was one integrated producer with 100% captive sales. The market was already in the highly concentrated range, and the joint venture would have significantly increased the HHI. Exxon and Shell together would have had more than 50% of the merchant sales in North America, but the facts indicated that coordinated interaction was a greater risk than a unilateral price increase. As in the other coordinated interaction cases, it was not likely that new entry would prevent coordinated behavior.
Unlike the refined petroleum products involved in Shell/Texaco and the hydrogen peroxide involved in Degussa, however, VI improvers are a differentiated product. Several different polymers, with different properties, can be used to make VI improvers, and there is ongoing development work to improve the product and meet changing standards. Both of those factors can make it more difficult to coordinate behavior, but we still had serious concerns that the firms could effectively coordinate. For example, instead of colluding on price, the firms could decide to allocate customers. Prices would increase just the same, and perhaps with greater certainty.
Several factors suggest that collusion would be successful. First, the demand for VI improvers is highly inelastic. It is needed to meet industry standards for motor oil, and there are no economic substitutes. In addition, it represents less than 10% of total cost of motor oil. Consequently, customers are insensitive to price, and suppliers have ample incentive to collude. Second, as one might expect when products are differentiated, there is some brand loyalty. Manufacturers of motor oil formulate and market their products based on the characteristics of a particular VI improver, and a change in formulation would require new testing and qualification to meet auto manufacturers' requirements. Thus, switching would entail significant costs and would take some time. Consequently, motor oil manufacturers would be reluctant to switch suppliers for only a small price increase.
Third, the proposed joint venture would have facilitated coordination by increasing the homogeneity among the firms with respect to their product mix. Along with VI improvers, motor oil producers also purchase other additives, including dispersant/inhibitor packages. Prior to the formation of the joint venture, one supplier of VI improvers did not offer DI packages, and one supplier of DI packages did not sell VI improvers. After the joint venture, all three of the merchant suppliers would sell both products and would have similar interests in maximizing the overall return from the full line of additives. Consequently, it would have been easier for those suppliers to coordinate their behavior, as well as more difficult for customers to shop around.
There are several lessons we can draw from the drug wholesaling cases, Shell/Texaco, Degussa/DuPont and Exxon/Royal Dutch Shell. First, the coordinated interaction theory is alive and well, and we will apply it in the appropriate case. Second, there is no standard checklist of essential facts that we apply to coordinated interaction cases. Some cases are more straightforward than others, but difficult cases don't get a free pass. Third, mergers resulting in a leading firm with a very large market share are not always challenged under a unilateral effects theory; we will allege coordinated effects when the facts make that appropriate.
Finally, the mere fact that products are differentiated, as in Exxon/Royal Dutch Shell, does not mean that a coordination theory will not apply. Differentiation can inhibit coordinated interaction to the extent that rivalry is multidimensional. Differentiation, however, inhibits coordination to a limited extent, for example, where there are only a small number of large firms that control the vast majority of market share. Coordination does not have to be perfect or complete, nor does it have to include all terms of competition. For example, inExxon/Royal Dutch Shell the concern was that the firms could avoid the complexities of coordinating prices by allocating markets. Further, differentiation might make coordination easier in two ways. First, even in a market with a large number of competitors differentiation may lead to sets of brands within the market that compete closely with one another, but not with other brands. If there is strong consumer preference for only one or two brands, only those firms producing the close substitutes need coordinate. Second, where differentiation is based on strong consumer preference, entry may be particularly difficult and costly, impeding the ability of rivals to take away share through entry or repositioning.
IV. Mergers in High Tech Markets
An increasing number of our merger investigations involve high technology markets. These investigations often pose complex and challenging issues in defining relevant markets, assessing competitive effects, gauging entry barriers, and determining the strength of efficiency claims.
One of the Commission's more prominent merger investigations during the last year involved the settlement of Digital's patent litigation with Intel, which resulted in cross-licensing of technology and the acquisition of a fabrication facility.(7) Digital and Intel are aggressive rivals in the present and future development of microprocessors, and Digital's Alpha microprocessor is a significant competitor both to Intel's Pentium microprocessor and to Intel's next generation IA-64 microprocessor. Digital sued Intel in May 1997 alleging that Intel's Pentium microprocessors infringed some of Digital's Alpha microprocessor patents. Intel countersued claiming that Digital's Alpha microprocessors infringed some of Intel's patents. Digital and Intel settled their suit in October 1997 by agreeing to broad patent cross-licenses, the sale of Digital's microprocessor production facilities to Intel, and an agreement that Intel would produce Alpha microprocessors for Digital. The sale of facilities to Intel did not include the intellectual property rights, design assets, or employees for the Alpha chip. Digital also agreed to design computer systems based on Intel's IA-64 architecture and chips.
The Commission's complaint alleged that certain aspects of the settlement were likely to reduce competition in three relevant markets: (1) the manufacture and sale of high-performance, general-purpose microprocessors capable of running Windows NT in native mode; (2) the manufacture and sale of all general-purpose microprocessors; and (3) the design and development of future generations of high-performance, general purpose microprocessors.(8) Antitrust concerns arose because in each of these markets Digital's Alpha chips happened to be the highest performance and most technologically advanced threat facing Intel's own microprocessors. While recognizing that the settlement would free Digital from the enormous cost of owning and operating its own chip fabrication facilities, the Commission was nonetheless concerned that Alpha would not remain competitively viable under the original terms of the Digital-Intel agreement. Intel would have the ability to interfere with Digital's supply of Alpha chips by withholding necessary cooperation at the manufacturing stage. Digital might also lack the ability and incentive to continue to actively develop and promote Alpha.
To resolve these concerns, Digital entered into a consent agreement under which it would license the Alpha architecture to Samsung and AMD or other suitable partners so that they would be able to produce and develop Alpha chips. Digital also agreed to begin the process of certifying IBM as a foundry for Alpha chips, thereby establishing a manufacturing alternative to Intel. Although typically we frown upon the use of non-structural relief, in this case the licensing arrangements seemed adequate, because of the strength, experience and market position of the licensees. Taken together, these provisions were intended to preserve Alpha as a viable product and a competitor to Intel's microprocessors.
In high-technology industries, as elsewhere, efficiency considerations are playing an ever increasing role in our analysis. Mergers that might have raised competitive concerns in previous years may now be approved when efficiencies are readily apparent. But efficiencies are most likely to make a difference in a "close" case, where the degree of competitive harm is relatively modest. Proponents of a transaction should not expect efficiencies to turn the ultimate result where the degree of harm is substantial. There are many cases, particularly those where the post-merger firm would have a significant market share, where claimed efficiencies are simply inadequate to overcome the anticompetitive impact of the merger.
In one recent investigation of a high technology merger, the parties argued that the transaction would achieve significant efficiencies from: (1) more effective competition against alternative technologies that might eventually challenge the parties' products at both the low and high ends of the market; (2) reduction in overlapping R & D; and (3) spreading fixed costs of R & D over greater number of customers.
The parties' first efficiency argument is one that I think we will hear with increasing frequency in high-technology industries. Because product performance and capability play such important roles in determining which firms achieve dominance in these industries, companies are legitimately concerned that they will be displaced either by next-generation products produced by competitors or by new technology. In this case, however, the argument was unavailing for several reasons. Available evidence suggested that, while the alternative technologies represented plausible long-term competitive threats, they would not attain a significant position in the market within two years. Under such circumstances, the parties' efficiencies claim was equivalent to a suggestion that competition in an existing market should be sacrificed so that the companies could more effectively compete in a different market several years in the future. Even if the parties were correct that their future competitiveness would have been enhanced, we will not consider arguments of this type where the price is sacrifice of competition in an existing market that is likely to continue to be of economic significance for a substantial period of time. I should add that the parties were also unable to show that continuing improvements on existing technology would not be sufficient to meet the competitive threat. In fact, the recent history of technological rivalry and innovation among the firms led staff to believe that a healthy market for products using their technology would persist for a significant period of time.
We also hear the parties' second efficiencies argument -- that combined R & D would eliminate unnecessary duplication -- quite often. That argument is not likely to succeed where innovation is a driving force behind competition. In this case, the claims were particularly unappealing because rivalry over successive innovations had produced lower prices and other significant consumer benefits over the years. We recognize that the reduction of R & D expenses may sometimes be a cognizable efficiency, but it is not when the result may be reduced competition.
The final efficiency justification is really just a statement that if you have a combined company with more customers but performing less R & D, the per unit cost of R & D will be less. In this context, the parties focused on enhancing their ability to compete in markets other than that in which they currently competed. Again, the claim could be characterized as an assertion that if they stop competing against one another they would be better able to compete against someone else. To be cognizable, an efficiency must offset the competition-reducing aspects of a transaction. A cost reduction of this type could be meaningful if it made the combined entity a better competitor within its existing market. In this case, however, the staff concluded that the reduction in R & D would lower overall R & D efforts in the existing market with a corresponding reduction in competition. The claimed "efficiencies" thus would result in no benefit to consumers.
V. Vertical Mergers
The 1984 Merger Guidelines describe several types of potential anticompetitive consequences of vertical mergers. All of these theories of harm are ultimately based on the merger's potential horizontal effects: competition is, after all, by definition horizontal. Broadly, there are two areas of concern identified in the Guidelines and the case law. The first is that vertical integration will raise entry barriers or foreclose non-integrated firms from a market in which the merged firm would operate. The second is that the integration may raise competitors' costs in an anticompetitive manner, or reduce the incentives of either the merged firm or its rivals to compete.
The barrier-to-entry and foreclosure concerns are essentially two sides of the same coin. If the newly integrated firm forecloses unintegrated rivals from raw materials on the upstream side or a market on the downstream side, the rivals will have to integrate themselves or perish, and new entrants will have to enter at both market levels in order to succeed. Of course, these effects could also be described as raising rivals' costs.
The potential reduction of incentives arises from access to competitively sensitive information. Because of its position in two levels of the market, the newly vertically integrated firm may relate to a rival both as a horizontal competitor and as a customer or supplier. In its position as customer or supplier, the merged firm may gain access to competitively sensitive information concerning its horizontal competitors. When a firm gains competitively sensitive information by participating in vertically related markets, it may be able to compete less aggressively. For example, if through its participation in an upstream market, the merged firm gains access to competitively sensitive information enabling it to reduce its uncertainty about a competitor's bids in a downstream market, the merged firm may be able to bid less aggressively in the downstream market. This concern extends to situations in which the competitor gains access to information about costs or technology with which it could estimate its rival's likely bid and adjust its own bid accordingly.
Further, integration may dampen the incentives of the non-integrated firm to compete. If the integrated competitor gets access to a non-integrated competitor's costs or technical information, that competitor's incentives to innovate or engage in research and development may be reduced. Where a firm knows that its competitors can "free-ride" on its innovations, the incentive to innovate may be seriously dampened. Similarly, if the non-integrated firm believes that it faces exclusion or discrimination from the integrated firm, it may choose to withdraw from the market or compete less aggressively.
Our recent vertical merger cases illustrate a number of these concerns.
PacifiCorp/The Energy Group PLC. In February, the Commission accepted for comment a consent order in a so-called "convergence merger" in the energy field. The "convergence" was between coal and electricity in the $10 billion proposed acquisition of The Energy Group PLC by PacifiCorp.(9) PacifiCorp provides retail electric service in seven western states: Oregon, Washington, California, Utah, Idaho, Wyoming, and Montana. The Energy Group owns Peabody Coal, which produces about 15 percent of the coal mined in the U.S., as well as a power marketer that trades electric power throughout the U.S.
A major competitive concern arising from this vertical merger was that PacifiCorp's control of two Peabody mines would enable it to raise its rival's generating costs and, consequently, the wholesale price of electricity in the western United States. It so happened that the Peabody mines were the only viable sources of coal for certain electricity generating plants that, at the margin, set the wholesale price of electricity in the region during certain generating periods. The merger would have frustrated the promise of new wholesale and retail competition brought about by the onset of deregulation in electricity markets. That is a concern we have in several markets undergoing deregulation -- telecommunications, cable television, and financial services, as well as electricity -- and we pay close attention to mergers in those industries. We do not want to see the extraordinary opportunity for new competition destroyed or diminished by anticompetitive alliances.
Consequently, the proposed order required the divestiture of the coal mines. We initially considered a settlement based largely on behavioral restraints. On further consideration, however, we believed this would be unsatisfactory -- in large part because we generally prefer structural relief that, once carried out, preserves competition through the operation of the market, not through ongoing "regulatory" policing by the Commission.
Merck/Medco. Merck and Co. acquired Medco Containment Services in 1993, as the first pharmaceutical manufacturer to integrate vertically into the then relatively new business of pharmacy benefit management. In 1995, the Commission took action against a similar acquisition, of PCS by Eli Lilly and Company, and pledged to monitor the industry carefully to determine whether further action against manufacturer-pharmacy benefit manager integrations was called for. On August 27, 1998, in accordance with that pledge, the Commission announced that it had accepted for comment an agreement with Merck to resolve antitrust concerns regarding the 1993 Medco acquisition.(10)
The complaint alleged that the merger tended to cause a reduction in competition for pharmaceutical products stemming from Medco favoritism toward Merck drugs in the "formularies" of drugs available under the programs it manages. In addition there were concerns raised because the merger had made it possible for Medco to provide Merck sensitive pricing information obtained from Merck's competitors, with the potential for fostering collusion among manufacturers. The complaint issued with the proposed order also discusses anticompetitive effects in eliminating Medco as an independent negotiator of pharmaceutical prices with manufacturers, and reducing other manufacturers' incentives to develop innovative pharmaceuticals.
The proposed Merck order would require Merck-Medco to maintain an "open formulary" -- a formulary including drugs selected and approved by an independent committee.(11) It also requires that Merck cause Medco to accept all discounts, rebates, or other concessions offered by other manufacturers on the open formulary, and to accurately reflect the impact of these factors on price in establishing relative rankings of products on that formulary. An additional order provision prohibits Merck and Medco from sharing "Non-Public Information" with each other, including information concerning other persons' bids, proposals, contracts, prices, rebates, discounts, and other terms of sale.
TRW/BDM International. Another recent Commission consent order, involving the acquisition by TRW Inc. of BDM International Inc., also dealt with vertical concerns.(12) TRW and BDM were each involved in the Defense Department's Ballistic Missile Defense program. A joint venture including TRW was one of two competitors bidding for a contract for a particular portion of this program. BDM was the Ballistic Missile Defense Organization's sole supplier of systems engineering and technical assistance -- or SETA -- services for that portion of the BDM program. In that capacity, BDM was responsible for developing technical and other specifications for this contract procurement, assessing bid proposals submitted by the contestants, and evaluating the cost and quality performance of the winning bidder. Under the proposed transaction, TRW would in effect have taken over these functions of BDM.
It is not hard to guess, then, that one of the Commission's competitive concerns had to do with what is colloquially called "the fox guarding the chicken coop" -- that the transaction as structured would have put TRW in a position to favor itself as a competitor for the contract, and, if it won, as to its fulfillment of the contract. A second concern was that as SETA contractor for the program, BDM necessarily came into possession of much highly sensitive information from both competitors for the procurement, and that access by TRW to such information from its competitor might enable TRW to raise its bid price and perhaps deprive its competitor of some of the fruits of its innovation. The remedy chosen in this case was thus a strong one, requiring TRW to divest BDM's SETA services contract with the BMDO and all associated assets.
Shell/Texaco.(13) This joint venture, which I discussed earlier, also had an interesting vertical aspect. Texaco owns the only heated pipeline that carries undiluted heavy crude oil from the San Joaquin Valley in California to refineries in the San Francisco Bay area, where Shell has a refinery that makes asphalt, as well as light petroleum products. Huntway Refining Company is the only other asphalt refiner in the Bay area. Both buy undiluted heavy crude from Texaco, transported by the pipeline, and refine it into asphalt (among other products). We were concerned because the transaction would allow the joint venture to raise Huntway's costs by increasing prices of undiluted heavy crude to Huntway relative to the price charged to Shell.
The consent order eliminates this risk by requiring the parties to enter into a 10-year supply agreement with Huntway, the terms of which must be approved by the Commission. The parties entered into such an agreement, which was made a confidential exhibit to the order. The order prohibits the joint venture from increasing the price or reducing the volume of crude oil supplied to Huntway, and also prohibits terminating the supply agreement except on terms identified in that agreement.
These four cases illustrate a mix of the theories of competitive harm that can arise from vertical mergers, as well as a variety of remedies to resolve those concerns. In view of the variation in remedies, you might ask whether there is a consistent framework that is being applied to these cases. I think there is. The key remedy issue in all of these cases is, what is the competitive harm and what will it take to fix the problem? As in the case of horizontal mergers, we are usually most comfortable with structural remedies because they provide the greatest assurance of an effective cure without a need for continuous monitoring. However, another important consideration is that vertical mergers, in general, have a potential for producing significant efficiencies that might be forsaken if we require structural relief. Thus, if it appears that a vertical merger will generate substantial merger-specific efficiencies, we will consider non-structural relief. There are some important qualifications to this, however. First, is there a form of non-structural relief that will effectively solve the problem? Second, or perhaps a corollary to the first, can private parties effectively protect themselves through contractual provisions? Third, is there a form of non-structural relief that will not be too regulatory?
Let me explore the resolution of those issues in the four vertical cases that I have discussed. In The Energy Group/PacifiCorp matter, a major substantive concern was the merged firm's ability to control the pricing of an important input to raise prices in the downstream market. The same concern was presented by the Shell/Texaco joint venture. There are three ways one might address that concern. Divestiture is one. Some form of price or rate control is another. A Commission-approved supply agreement between purchaser and supplier is another. An order that would impose a price or rate control is something we rarely, if ever, would find acceptable; almost by definition, it is too regulatory. That leaves the other two options. In The Energy Group/PacificCorp we opted for divestiture, while inShell/Texaco the order addressed the problem through a long-term supply agreement. Why the different approaches? As I noted earlier, a critical issue is whether the remedy effectively solves the problem. A divestiture does that with certainty, but if the parties can offer an effective alternative, we will consider it. But we would have to satisfy ourselves that the alternative was negotiated by sophisticated parties, that it could be monitored and enforced without undue difficulty, and that it would address our concerns -- including competitive harm in the downstream market -- and not just the concerns of the downstream competitor. Those are stringent requirements and will be met only in rare cases. The Shell/Texaco consent order is one of the few cases where such an arrangement is satisfactory. The supply agreement generally preserves a business arrangement that had arisen before the merger (the pipeline supply of heavy crude oil to Shell's competitor), so preserving that relationship -- which is what the order essentially does -- preserves the competitive market conditions that existed before the merger.
Now consider the other two cases. Both Merck/Medco and TRW/BDM raised concerns about exclusion from the relevant market and the anticompetitive use of competitively sensitive information. The proposed order in Merck/Medco contains, among other provisions, an information "firewall" to prevent the misuse of competitively sensitive information. In TRW/BDM, however, we required divestiture. Why the difference? Because TRW/BDM presented another competitive problem that could not be resolved effectively without divestiture. The concern was that the SETA contractor (BDM) set specifications for the defense contract, evaluated bids on behalf of the customer and monitored performance. Unless the SETA contract were divested, TRW's acquisition of BDM would have placed TRW in the position of both a competing bidder for the defense contract and the evaluator of bids and performance.
In contrast, Merck/Medco also involved competitive concerns that went beyond the potential misuse of confidential information, but they could be dealt with through non-regulatory injunctive provisions. For example, to prevent Medco from discriminating against Merck's competitors, the consent order requires the creation of an independent formulary. An independent Pharmacy and Therapeutics ("P&T") Committee selects the drugs to be placed on, or removed from that drug formulary. This required minimal government involvement in the operation of the formulary, because the decisions of which drugs would be included would be made by the independent P&T committee.
In sum, vertical mergers require a careful assessment of potential remedies as well as the competitive problems they are intended to address. There is no "one size fits all" solution for these cases. While that detracts to some extent from predictability, we also benefit from an efficient solution.
Endnotes:
1. See Statement of the Federal Trade Commission Concerning Mergers and Corporate Consolidation in the New Economy, presented by Robert Pitofsky, Chairman, before the Committee on the Judiciary, United States Senate, June 16, 1998.
2. See Pepsico, Inc. v. Coca-Cola Co., 1998 U.S. Dist. LEXIS 13440 (S.D.N.Y. Aug. 27, 1998) (relying on the Staples and Cardinal decisions (that a channel of distribution can define a market, notwithstanding the fact the products may be available from other sources) to hold that Pepsi's alleged relevant market of "sales of fountain-dispensed soft drinks distributed through independent food service distributors" was sufficient to survive a 12(b)6 motion).
3. Shell Oil Co., FTC Dkt. No. C-3803 (April 21, 1998) (consent order).
4. Degussa Aktiengesellschaft, FTC Dkt. No. C-3813 (June 19, 1998) (consent order).
5. Exxon Corp., FTC File No. 9710007 (consent agreement accepted for public comment, Aug. 20, 1998).
6. Using section 7 to challenge a merger where there is a history of oligopolistic behavior is particularly appropriate. As the Supreme Court stated in Brooke Group, Section 7 seeks to prohibit "excessive concentration, and the oligopolistic price coordination it portends." Brooke Group v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 229-30 (1993). Professor Areeda observed that such oligopolistic pricing
is feared by antitrust policy even more than express collusion, for tacit coordination, even when observed, cannot easily be controlled directly by the antitrust laws. It is a central object of merger policy to obstruct the creation or reinforcement by merger of such oligopolistic market structures in which tacit coordination can occur.
Phillip Areeda & Herbert Hovenkamp, Antitrust Law, ¶ 901'a, at 489 (1997 Supp.).
7. In re Digital Equipment Corp., C-3818 (July 14, 1998)(consent order).
8. In re Digital Equipment Corp., Complaint at ¶¶ 11-13.
9. PacifiCorp, FTC File No. 9710091 (consent order accepted for comment Feb. 18, 1998). The proposed order was withdrawn before becoming final, and the investigation closed by the Commission, because during the public comment period PacifiCorp withdrew its bid for The Energy Group and eventually sold virtually all of its Energy Group stock.
10. Merck and Co., Inc., File No. 9510097 (consent agreement accepted for comment Aug. 27, 1998).
11. Under the order, Medco would be free to offer a more restrictive formulary as well as an open formulary.
12. TRW Inc., FTC Dkt. C-3790 (consent order made final Apr. 6, 1998).
13. Shell Oil Co., FTC Dkt. C-3803 (consent order made final April 21, 1998).