Rediscovering Coordinated Effects

Good afternoon. It is a pleasure to be here, and to be able to share the stage once again with my good friend, FTC Chairman Tim Muris. It has been said that life is largely a matter of expectation. Throughout my career in the antitrust field, I have been first and foremost a merger lawyer, and as a result I expected that the majority of my time at the Antitrust Division would be devoted to merger matters. We were, after all, in the midst of an unprecedented wave of corporate mergers and acquisitions, which already had begun to slow but was still producing a steady stream of transactions. Having been at the Division a little over a year now, I have found that in life expectations do not always match outcomes.

Just two weeks after arriving at the Division, the D.C. Court of Appeals issued its decision in the Microsoft case. Microsoft and a number of other non-merger matters, including our efforts to increase international antitrust harmonization, have in fact occupied a great deal of my time during my first year in office. The downturn in the stock market, the tragic events of September 11, and the spate of corporate fraud scandals, in turn, each has contributed to reduced levels of merger activity over the past year.

Merger activity is significantly down compared to prior years. We have received notification of just 1,020 Hart-Scott-Rodino transactions for Fiscal Year 2002 to date, down from 2,075 for the same period of Fiscal Year 2001. My immediate predecessors had the benefit of 4,076 HSR transactions for the same portion of Fiscal Year 2000. Even taking into account last year's increase in the HSR filing threshold, 2002 merger figures are still below previous levels of activity, reflecting the well-known slow down in corporate transactions. The telecommunications and technology sectors have been particularly slow, and those areas of the economy typically supply some of our more significant transactions.

Notwithstanding the general slowdown, the Division has investigated a number of significant transactions over the past year. These matters include:

• General Dynamics/Newport News. General Dynamics and Newport News were the only two nuclear-capable shipyards and the only designers and producers of nuclear submarines for the U.S. Navy. The two shipbuilders also lead opposing teams to develop the next generation propulsion system for use in submarines and surface combatants, so-called electric drive. The Department of Defense ("DoD"), the only customer, opposed the transaction on competitive grounds. The complaint alleged that the combination would create a monopoly in nuclear submarines design and construction, and would substantially lessen competition for electric drive and surface combatants. After the parties terminated their merger agreement, Newport News received a second bid from Northrop Grumman, which did not raise significant competitive issues.
• Suiza/Dean. Suiza and Dean were dominant firms in several geographic markets for fluid milk processing and school milk markets. The parties agreed to divest eleven dairies to National Dairy Holdings, L.P. (NDH), a newly formed partnership that is 50 percent owned by Dairy Farmers of America Inc. (DFA), a dairy farmer cooperative. The parties also agreed to modify Suiza's supply contract with DFA to ensure that dairies owned by the merged firm in the areas affected by the divestitures would be free to buy their milk from sources other than DFA.
• United/USAirways. At the time of the transaction, United and US Airways were the second and sixth largest U.S. airlines. The Division concluded that US Airways is United's most significant competitor on densely-traveled, high revenue routes between their hubs, such as Philadelphia and Denver, as well as for nonstop travel to and from Washington D.C. and Baltimore, and on many routes up and down the East Coast. The acquisition would have given United a monopoly or duopoly on nonstop service on over 30 routes, where consumers spend over $1.6 billion annually, and substantially limited the competition it faces on numerous other routes representing over $4 billion in revenues. The parties abandoned the transaction after the Division indicated its intention to challenge it.

In fact, since June 2001, the Division has successfully challenged 18 of the 19 transactions it had deemed anticompetitive. The Division was unsuccessful in seeking to block the Sungard/Comdisco merger, a transaction the Division asserted was likely substantially to lessen competition in the market for shared hotsite disaster recovery services. Five of these matters were resolved by consent decree, seven through a "fix-it-first" restructuring, five were abandoned after the Division indicated that it would file suit to challenge the transactions, and one -- General Dynamics/Newport News -- was abandoned after the Division filed suit.

In addition, this past April, we announced a proposed settlement of a gun-jumping complaint against Computer Associates and Platinum Technology, alleging violations of both Section 7A of the Clayton Act and the Sherman Act. In another Section 7A case, the Hearst Corporation and its parent, The Hearst Trust, agreed to pay a civil penalty of $4 million to resolve charges that the company failed to produce key documents required to have been supplied along with its pre-merger notification of the acquisition of Medi-Span. The Division filed the lawsuit at the request of the Federal Trade Commission, and the civil penalty is the largest a company has ever paid for violating antitrust pre-merger requirements.

Several dozen merger investigations are in the queue, including DirecTV/Echostar, AT&T/Comcast, Univision/Hispanic Broadcasting and TRW/Northrop Grumman.

Our plate also has been full with civil non-merger matters, some of them involving the resolution of cases brought by the prior Administration. Of course, the big news this past year in the non-merger area has been our settlement of the Microsoft case. The settlement, together with hearings on separate remedial proposals being advanced by a group of dissident states, is under review by the Court. Additionally, appeals are underway in the American Airlines predatory conduct case and the Visa/Mastercard case. Since my arrival at the Division, we have launched a number of important joint venture investigations involving, among other things, on-line media, financial services and electronic air passenger ticketing. Joint ventures are a high priority for the Division, in part because we believe that many firms are turning to joint ventures as an alternative to full-out mergers, and in part because joint ventures are an important way in which competitors interact with each other in emerging markets.

Our criminal enforcement program is moving along at a swift pace. The Division currently has 99 pending grand jury investigations, approximately 35 of which implicate suspected international cartel activity. Over the course of the past year the Division also has obtained several record-breaking jail sentences against participants in international and domestic price-fixing and bid-rigging conspiracies. In January 2002, Sonny Shelton, a former Guam government official, was sentenced to 10 years in jail -- the longest antitrust-related jail sentence ever imposed -- for orchestrating a bid-rigging, bribery, and money laundering scheme involving FEMA-funded contracts in Guam. Also notable was the prosecution and conviction of A. Alfred Taubman, the former chairman of the board of Sotheby's Holdings Inc., who was sentenced to serve one year and one day in prison for his role in a price-fixing scheme between Sotheby's auction house and Christie's.

Returning to the merger area, we have tried to use our time and merger enforcement resources productively -- sharpening our skills and techniques for when intensive merger activity resumes. As Dwight Eisenhower once said, "In preparing for battle I have always found that plans are useless, but planning is indispensable." To that end, the Division has undertaken several efforts designed to improve our ability to investigate mergers effectively and efficiently. These efforts include our modernization effort; the Merger Review Process Initiative; the aborted Clearance Agreement with the Federal Trade Commission; our ongoing study of merger remedies; the development of a best practices manual on merger investigations; and various international harmonization initiatives.

The modernization effort implemented earlier this year is the first congressionally approved reorganization and modernization of the Division in more than two decades. It is designed to improve our effectiveness as enforcers by concentrating industry expertise within particular sections of the Division and giving those sections broad enforcement responsibility for both civil merger and non-merger matters. Similarly, we implemented the Merger Review Process Initiative in order to more quickly identify the critical legal and economic issues raised by transactions, facilitate more focused investigative discovery, and provide for an orderly process for the evaluation of evidence. The success of the Initiative depends in large part on the willingness of our lawyers and the parties to work together in identifying at the earliest possible point the issues that require ongoing investigation, and work together to develop a sensible plan for investigating these issues. We have received some very positive feedback thus far regarding the Initiative, but the true test will come when we have a broader array of mergers under review.

On a related track, we have had an ongoing project to evaluate and institutionalize "best practices" for conducting merger investigations. This project, which has been under the direction of our Director of Operations, Connie Robinson, and has involved active consultation with the Federal Trade Commission, will result in the creation of a best practices manual that will be a valuable tool for our staffs. The type of "self-benchmarking" reflected in the best practices initiative will allow us to embrace merger techniques that have proven successful in real investigations, while eliminating techniques that have proven unproductive. We look forward to working with the private bar, as we constantly attempt to improve our performance.

Our landmark but ill-fated clearance agreement with the FTC also was part of our process of procedural reform in the merger area. By reducing the time spent in clearance disputes, more time could be spent on actual investigations during the first HSR waiting period. During its brief life, the agreement had reduced clearance to a one-day process, and for the first time in decades eliminated all pending clearance disputes. The agreement itself did not change or alter substantive antitrust enforcement, nor did it transfer industry responsibility from one agency to the other. It merely institutionalized in advance the results that should have been dictated by the pre-existing, experience-based system. Much to our disappointment, the agreement had to be voided in May of this year due to the threat of budgetary reprisals against the agencies. Chairman Muris and I, however, remain committed to improving the efficiency and effectiveness of the clearance process. Nevertheless, despite our best efforts, in recent weeks, the average time for clearing matters to the agencies has risen to five days.

On the international front, our divergence with the EU last year on conglomerate merger theory and the role of efficiencies in merger enforcement signaled a compelling need for a more substantive dialogue with our EU counterparts. This divergence has been an impetus for a much higher level of substantive consultation and exchange in the merger area. It is vital that we continue to engage in such a dialogue, for I believe that the EU is at a critical juncture with respect to merger policy and enforcement, with the management changes underway within DG-Comp, the recent AirTours decision, pending appeals in a number of important merger cases, and the EU having announced its intention to develop its own merger guidelines.

Fortunately, we continue to make excellent progress in our EU/US Merger Working Group, and have ongoing projects designed to achieve greater substantive convergence on competitive effects analysis and the role of efficiencies. Over the coming weeks we also hope to reach final agreement with the EU on a set of procedural best practices for coordinating merger investigations. Again, these efforts, are designed to make our enforcement efforts more effective and more efficient.

Complementing these efforts is our work with the International Competition Network (ICN). In October 2001, the U.S., the EU and 14 other jurisdictions established ICN as a new vehicle through which antitrust agencies from developed and developing countries will seek to develop consensus positions on specific proposals for procedural and substantive convergence in antitrust enforcement. ICN has initiated two major projects during its first year of existence. The first one is the development of best practice recommendations in the area of competition advocacy. The second is the development of best practices for merger review. 60 jurisdictions on six continents have joined the ICN, and we look forward to our first meeting of ICN members this September in Italy. Based upon the work performed so far, I think the antitrust community will be pleasantly surprised about what the ICN can accomplish.

Today, I would like to talk about another important element of merger review -- competitive effects. The 1992 Horizontal Merger Guidelines amplified the significance of competitive effects analysis, identifying it for the first time as one of the necessary elements of merger review under the Guidelines, and explicitly identifying unilateral effects and coordinated effects as alternate theories of competitive harm. It is probably fair to say that, prior to the 1992 Guidelines, anticompetitive effects -- in particular, an increased propensity for tacit collusion -- were largely presumed from high levels of concentration in combination with high barriers to entry.

Despite the fact that much of the case law since the introduction of the structural presumption in Philadelphia National Bank has focused on coordinated effects or collusion, one interesting side-effect of the 1992 Guidelines has been the emergence of unilateral effects as the predominant theory of economic harm pursued in government merger investigations and challenges.

Everything in the antitrust world is complex. Thus, it would be incorrect to attribute this phenomenon to any single factor. The following factors have, in my view, been significant contributors:

• First, there is the growing economic focus on strategic behavior and the belief by many economists that unilateral price changes resulting from mergers can be predicted and demonstrated empirically. There is no corresponding set of predictive tools for coordinated effects.
• Second, coordinated effects analysis, relying as it does upon a complex set of market variables that may point in opposite directions, can be perceived to be less determinate than the types of effects predicted under unilateral theory.
• Third, even once all of the factors have been analyzed, we have yet to develop any well-accepted science that specifies the precise level of market concentration or the minimum number of competitors at which coordination is likely.
• Fourth, the analysis set forth in the Guidelines does not do as much as some might like to determine why maintaining the independence of one of the specific parties to the proposed transaction is an important enough constraint upon coordinated interaction to justify a challenge.

Within the agencies, unilateral effects cases are perceived by many to be easier to analyze and more likely to gain economic support. Outside of the agencies, however, unilateral effects cases are no bargain. The case law of Section 7 largely has developed on a coordinated effects model, and courts appear to be more comfortable addressing merger issues based upon the traditional approach. Moreover, if placed under sodium pentathol, most economists would concede that market definition is not particularly important in unilateral effects analysis, an additional fact that can make it difficult to square the case theory with the way many courts tend to view merger issues. An additional complicating factor in unilateral effects litigation is the heavy dependence upon theoretic economic testimony -- particularly econometrics and merger models. Economic testimony, of course, is inherently easy to attack, especially because each side will have its experts advocating opposite conclusions. Finally, very often, the effects predicted by unilateral theory are small in percentage terms.

The foregoing discussion is not intended to suggest that unilateral effects cases are not important or should not be brought. Where the predicate facts are present and substantial adverse effects are clear, enforcement action is appropriate. Rather, it suggests the need for the agencies to sharpen their analytical tools with regard to coordinated effects. In other words, unilateral effects should not be the theory of choice simply by default. If we reach too quickly for unilateral effects theories to the exclusion of meaningful coordinated effects analysis, we might miss important cases that should be brought or craft our relief too narrowly in cases that we actually pursue.

For the last several months, I have had a team of lawyers and economists led by the Division's Deputy Assistant Attorney General for Economics, Michael Katz, and Acting Chief of our Competition Policy section, Andrew Dick, looking closely at coordinated effects analysis. I am very encouraged by their work thus far. Phase I of this project has involved a comprehensive review of the economic literature, case law and enforcement practice with respect to the three elements of coordinated interaction set forth in the 1992 Guidelines, with a view to improving our focus on the correct economic principles and evidentiary issues.

The Guidelines provide that in order for firms to coordinate their business practices, they must be able to do three things: 1) reach terms of coordination that are profitable to the firms involved, 2) detect deviations that would undermine coordination, and 3) punish any such cheating that occurs. There are many different market conditions that can facilitate coordinated interaction. In the context of merger review, of course, a fundamental issue is how one or more market conditions could interact with a change in market structure to increase the likelihood and success of post-merger coordinated interaction.

This exercise has been a useful one. Among other things, it has been helpful for us to examine more closely how market factors, such as product homogeneity and the size or frequency of purchases, affect the environment for coordination. Additionally, it has been useful to examine the circumstances as to whether and when a large buyer might be able to undermine coordination. Revisiting the various factors that potentially might affect coordination, and providing additional insight into how such factors might interact with each other, will assist our staffs in identifying solid coordinated effects cases and allow us to more readily develop the evidence we need to prove them in court.

Simply sharpening our focus on market characteristics, however, will not, by itself, instill greater confidence in coordinated effects analysis. The age-old questions of merger policy still persist. When exactly do market conditions become ripe for coordination and why does this particular merger matter in creating such conditions? Is it enough to demonstrate that the merger will eliminate important constraints upon coordination, or must the government go much further to present a much more detailed story as to how some specific form of post-merger coordination is more likely to occur? What must the government prove and to what level of certainty?

Phase II of our project, then, is designed to improve our ability to answer at least one of these questions: why does this particular merger matter. This will involve a much more detailed focus on the specific characteristics of the parties to the transaction and their roles as competitors in the market, pre-merger: the nature of their products, their cost structures, past competitive strategies, etc. Per the Guidelines, we also will focus on the ways in which the merger, including any efficiencies likely to result therefrom, will alter their incentives. In this way, we will be able to focus more directly on how individual firms affect the competitive dynamic in a market and how their elimination through merger may impact upon the likelihood of coordination.

This type of analysis is suggested in the section of the Guidelines addressing the elimination of competitive mavericks. The Guidelines, however, should not be interpreted to suggest a search for a single maverick, but rather an analysis of the characteristics of particular firms in terms of their incentives and propensities to form or undermine tacit or explicit coordination. Where a firm's presence in the market acts as an important constraint upon coordination, its elimination through merger is cause for enforcement concern. Needless to say, our analysis of "maverickness" will have to account for the likely reactions of other firms in the industry and the manner in which the merger may affect their tendencies toward coordination. Developing a more fully-formed view of the competitive significance of the parties proposing to merge, however, can only improve the quality of our merger analysis and increase our chances of prevailing when we challenge transactions based upon coordinated effects theories.

Throughout this process of rediscovering coordinated effects, we will continue to draw upon the prevailing economic literature, case precedent, and case experience, as well as share our perspectives with our colleagues at the FTC, who are undertaking a similar endeavor. We hope that as a result of our efforts we will be better analysts and advocates, and that we will renew the focus on coordinated effects analysis.

I look forward to reporting upon our progress over the coming months. Thank you for your time and attention this afternoon.