Skip to main content


The 1996 Annual Meeting, Section of Antitrust Law, The American Bar Association
Orlando, Florida
Susan DeSanti, Former Deputy General Counsel for Policy Studies


I am pleased to be on this panel to discuss new developments and directions in antitrust law and policy. This topic is particularly relevant to my job as Director of Policy Planning at the FTC. Last fall, Policy Planning organized hearings on whether marketplace changes in competition require any new directions or developments in antitrust law and policy. Following those hearings, Policy Planning staff undertook a thorough review of the issues that were addressed in the hearings and prepared a report, publicly released this June, on “Anticipating the 21st Century: Competition Policy in the New High-Tech, Global Marketplace.” This report contains some proposals for adjustments in antitrust policy.

Today I would like to focus on one of the proposals that has received a good deal of attention: the treatment of efficiencies in merger analysis. I will explain the efficiencies proposal and then discuss in general terms how it would work, using two examples that were developed from combining and altering the facts of a few actual mergers. As is customary, my remarks reflect only my views and do not necessarily represent the views of the Commission or any Commissioner.(1)

The FTC Hearings and Staff Report

Last fall the FTC held two months of public hearings on how increasing globalization and innovation are affecting competition and what, if anything, this agency should do to keep pace with new developments. The hearings were based on the FTC’s power to investigate and spotlight marketplace developments.(2) This investigatory mission was part of the FTC’s original mandate and has been the source of important studies in the past. For example, a study of the radio broadcasting industry influenced passage of the Radio Act of 1927, a predecessor to the Federal Communications Act of 1934. Similarly, the FTC’s disclosure of securities abuses played a role in heightening Congress’ recognition of the need for securities industry regulation, which led to the Securities Act of 1933.

More than two hundred witnesses participated in the FTC hearings last fall. From the thoughtful and valuable testimony of many business and consumer representatives, lawyers, economists and academics, there emerged several clear market realities and some corresponding new directions for competition policy. Some of the most prominent themes included the following. First, competition today is indeed global, and international trade and investment is expanding at an increasingly rapid rate. Second, the very nature of competition is changing. Competition today not only includes the price at which a product is sold but the ingenuity, variety, and speed of development of new goods and services. Third, competitor collaborations are increasingly important for staying at the forefront of markets that are global or characterized by innovation-based competition. These new market realities affirm the need for correspondingly up-to-date competition and consumer protection enforcement policies.

Based upon the testimony from the hearings last fall, as well as case law, legal and economic literature, and discussions with FTC staff, the Policy Planning staff issued its report in June. Volume One of the report focuses on competition policy. It summarizes the hearings testimony, and then analyzes and makes recommendations on several topics including: joint ventures, innovation markets and innovation competition, networks and standards, geographic market definition, and efficiencies. The report truly represents a collaborative effort among the Policy Planning staff, and it incorporates comments and suggestions from many others as well. For today, I will focus my remarks on efficiencies.

Proposed Treatment Of Efficiencies In Merger Analysis

Virtually every witness agreed that mergers sometimes may lead to substantial efficiencies. Because of this, many thought that efficiencies should play an enhanced role in merger analysis, and that the agencies should reassess the treatment of efficiencies under the 1992 Merger Guidelines. Support for this began building from the very first day, when James Rill, a principal architect of the current Guidelines and former Assistant Attorney General, urged that “the time was ripe” to address and more fully articulate how merger analysis should handle efficiencies.

The staff report proposes that merger analysis ask how any credible, merger-related efficiencies are likely to affect the post-merger competitive dynamics of a relevant market. This represents only one possible conceptual framework for analyzing efficiencies claims in mergers; others exist. Indeed, based on the report’s recommendation, the FTC and DOJ have established a joint task force to consider whether and, if so, how to revise the efficiencies section of the current Guidelines.

But I would like to share more of Policy Planning’s thinking on why the staff report’s efficiencies proposal makes sense. The analytical framework in the report views competition as the means by which Section 7 of the Clayton Act ensures that consumers receive the best quality and choice of products at the lowest prices. This may appear to be a simplistic statement. But what if merger analysis simply asked whether post-merger prices would be lower or higher? As Oliver Williamson taught us, even a monopolist may be motivated to lower price if it is able to achieve substantial efficiencies.(3) So, if merger analysis asked only whether a merger might result in lower prices, even a merger to monopoly might be justified, depending on the magnitude of the efficiencies available through the merger and the monopolist’s demand curve. But where would that merger to monopoly leave consumers in the long run, with no competition to ensure incentives to pass on to consumers subsequently obtained efficiencies or to improve product quality, for example? To protect consumers from a monopoly, regulation is typically in order. But, given what we have learned about the disadvantages of regulation and the relative paucity of true natural monopolies, it would seem a rare case indeed in which an antitrust enforcement agency would permit a merger to monopoly and suggest regulation of the new monopolist to protect consumers over the long run.

To protect consumers over the long run, there must be competition. Thus, the crucial issue is the level of competition post-merger; it must be sufficient not to substantially lessen competition below what it would have been absent the merger. Therefore, under the efficiencies proposal in the FTC Staff Report, the significance of efficiencies would lie in how they likely would contribute to post-merger competition. Within such a framework, it is important to take a dynamic, rather than static, view of competition. Under a dynamic view, our approach recognizes the competitive significance of efficiencies “over time.” It looks at how merger- related efficiencies likely will affect market competition within a reasonable period of time.

This approach obviously assumes that efficiencies, in some circumstances, may affect post-merger competitive dynamics. Although this is certainly not always true, it does appear that mergers sometimes generate efficiencies, and the effects of those efficiencies may be procompetitive rather than anticompetitive. In other words, if a merger is likely to achieve efficiencies, those efficiencies may affect the merged firm’s abilities and incentives in ways that deter any increased likelihood of the exercise of market power post-merger, or even make the market more competitive. Efficiencies likely to be obtained through a merger may increase the competitiveness of the merged firm and improve (or not impair) the competitive performance of the market(s) in which it operates, ultimately resulting in lower prices, increased output, and/or higher quality goods or services for consumers and other buyers. Such an efficiencies justification would enable credible efficiencies to be evaluated for their contribution to the overall likely competitive effect of the merger in a relevant market.

Under this “competitive dynamics” framework, the proposed efficiencies analysis would involve a two-step inquiry: (1) is the merger likely to result in credible efficiencies, and (2) if so, how are those efficiencies likely to change the merged firm’s abilities and incentives so as to deter the likelihood of lessened competition or increase competition in the relevant market post- merger? Under this proposal, it is not enough for the merging parties simply to allege that the merger will produce credible, substantial efficiencies. The parties must take those demonstrated, credible efficiencies and show how they likely would make the relevant market not less, or even more, competitive than it otherwise likely would be without the merger. In other words, efficiencies that likely would benefit the acquirer, but not competition, would not be recognized. At the end of the day, this proposal squarely affirms, and indeed relies upon, the principle that competition in the market is the best method for protecting consumers.

Before moving on to the examples using the proposed efficiencies approach, let me address some of the most frequently asked questions about the proposal.

What types of efficiencies are relevant? The proposal suggests that the FTC consider a potentially wide range of efficiencies (both product and process), from economies of scale and plant specialization to distributional, promotional, transactional, managerial and innovation efficiencies. However, not all efficiencies are created equal. Not all are equally susceptible to reliable proof, nor are all efficiencies equally likely to enhance competitive dynamics. For example, plant and production economies of scale are generally accepted as important to a firm’s ability to compete and are often subject to reasonable assessment. Innovation efficiencies may make an important contribution to competitive dynamics and the country’s overall welfare, but may be much more difficult to prove and to assess. Other cost savings also may be competitively relevant, but it may be difficult to assess whether the proposed merger actually would be likely to achieve them. Still other cost savings may be relatively straightforward, but not appropriately categorized as competitively relevant. As just one small example, how should an agency assess the competitive relevance of cost savings to be achieved because the merged firm will stop giving away free samples to customers? Do such cost savings represent procompetitive efficiencies, or do they represent an anticompetitive effect of the proposed merger, in that two firms that previously competed by giving away free samples now propose to eliminate that element of competition between them? An agency might never need to answer that question, if it appears that each firm likely would achieve such a cost savings through unilateral action. Nevertheless, we recognize that the efficiencies proposal in the FTC staff report simply provides an analytical framework for thinking about such questions; it does not provide automatic answers for specific questions, especially where the answers may depend on an overall assessment of the facts of a particular case. The report proposes that the FTC remain open to consideration of reasonable and demonstrable claims. At the end of the day, the weight and significance accorded to different types of efficiencies should be a function of their magnitude and probability, the degree to which they likely will enable the merged firm not simply to be a stronger competitor but to enhance (or not lessen) competition and thus benefit consumers, and the delay with which these consumer benefits are likely to be realized.

Who bears the burden of producing evidence of efficiencies? The proposal encourages the merging parties to make their efficiency submissions (including supporting documents) to an agency at an early stage of its review of the transaction. This would facilitate an accurate and serious assessment of the nature, probability, and magnitude of the claimed efficiencies and their likely effect on the competitive performance of the relevant market.

Under the staff proposal, however, efficiencies would not be part of the government’s prima facie case. Rather, the parties would bear the burden of producing evidence of competitively relevant efficiencies in seeking to rebut a presumption of likely anticompetitive effect. Given the information disparity between the agency and the parties with respect to efficiency claims, this burden of production is entirely reasonable. It also is generally consistent with existing merger law.(4) For several reasons, the proposal also concludes that the parties’ efficiencies evidence should not be subject to a “clear and convincing” standard. First, efficiencies evidence to rebut a showing of likely anticompetitive effect should not be held to a higher standard of proof than the elements of the case demonstrating a likely anticompetitive effect. Second, “clear and convincing” is technically a burden of persuasion standard, and we fully recognize that the burden of persuasion as to whether a transaction is likely to lessen competition remains with the government.

Must the efficiencies be merger specific? Phrased differently, must the proposed merger be the least restrictive way of achieving the claimed efficiencies? To answer this question correctly, we must understand why we ask it. We thought it important to step back and ask what Congress meant when it said that Section 7 was intended to arrest anticompetitive tendencies in their incipiency. We found that courts have contrasted a likely future with the merger to a likely future without the merger to assess a merger’s likely competitive effects. The critical focus, therefore, is what is likely to happen if the merger occurs as compared to what is likely to happen if the merger does not occur.

This should logically be the test for determining which efficiencies are cognizable as well. Under our proposal, the agency should not consider procompetitive efficiencies that likely would occur even without the proposed merger. For example, assume that one of the two firms likely would achieve equivalent efficiencies unilaterally. Such efficiencies should not be considered as procompetitive benefits of the merger, since they would be part of the probable future without the merger. Similarly, if the parties could obtain comparable efficiencies through licensing and, without exception, the industry practice is to do so, such efficiencies likely would occur without the merger and should not be considered a procompetitive benefit of the transaction. Thus, our proposal would reject efficiency claims if there were a significantly less restrictive means of achieving comparable efficiencies, and it would be practicable and feasible as a business matter to do so.

Must efficiencies be passed on to consumers? To evaluate efficiencies in terms of their competitive effects necessarily subsumes the question of the extent to which benefits from the efficiencies must be passed on to consumers. If the likely efficiencies from the merger will result in a more competitive market, or prevent a lessening of competition within the market, then the post-merger market itself will retain sufficient competition to ensure that, over time, merger- generated cost savings will benefit consumers through lower prices or improved quality goods.(5)

Examples Of The Proposed Approach To Analyzing Efficiencies

The Staff Report. The report suggests several ways in which claimed efficiencies might change the merged firm’s abilities and incentives so as to deter the likelihood of lessened competition post-merger or increase competition in the relevant market. For example, if merger- related efficiencies would enable a firm to lower its costs, those lowered costs may disrupt market conditions so as to make coordinated interaction less likely or to disturb the terms by which firms previously were able to coordinate their conduct. Similarly, if a merger combined complementary technologies and thus enabled the creation of a new or improved product, the increased product variety, in itself of value, might stimulate competition or impede competitors’ ability to coordinate. Likewise, if merger-related efficiencies eliminated a technology disadvantage, the merged firm might become a more significant constraint on market leaders. And merger-related efficiencies might enable the merged firm to reposition itself and constrain existing unilateral price elevation in a market for differentiated products.

Beyond these general examples of how merger-related efficiencies potentially could positively affect competition in a post-merger market, it perhaps will be helpful to demonstrate the proposed efficiencies approach by analyzing a couple of hypothetical mergers. In order to lend some degree of “real world” quality, the following merger examples are loosely based on combinations of facts from a few actual cases that have come before the agency. Industries, parties, and some market conditions necessarily have been altered to maintain confidentialities.

Assessing merger-related efficiencies, of course, is only one of several factors in merger analysis that, together, determine the overall likelihood that a merger will substantially lessen competition. Although I will touch on some of the other factors -- market definition, anticompetitive effects, entry -- my focus today will be primarily on efficiencies. Nevertheless, it is important to keep in mind that a judgment about the likely overall competitive effects of any merger necessarily will depend on putting together all of the factors in merger analysis.

Example #1 -- High-Tech Industry. There are currently four companies that produce high-end computer data processing software for a particular commercial use: Companies A and B, each with a 40% market share, and Companies C and D, each with a 10% market share. Companies C and D propose to merge. The relevant product market for the proposed merger is high-end computer data processing software. Developed for specific commercial use, these programs provide functional sophistications and creative capabilities not found in low-end data processing software programs. There are no substitutes at the 5 percent price test level. The geographic market for the industry is worldwide. All companies in this product market sell their software globally. The merger would reduce the number of players in this highly concentrated market from four to three and would significantly increase market concentration. Although the resulting market share of 20% for merged C/D is insufficient to raise concern about possible unilateral anticompetitive effects, the reduction from four to three players might raise concerns about the possibility of coordinated interaction.

Entry into the market would not be timely, likely or sufficient to deter or counteract anticompetitive effects of the merger. Development of just a prototype software package would likely take one to two years and entail large sunk costs. It is estimated that it would take three years to develop a marketable product and would require a large and relatively high-risk commitment in an industry with total sales of under $100 million.

What is the likely effect of this proposed merger on competition in the relevant market? C and D merging would combine the number 3 and number 4 producers of high-end computer data processing programs for commercial use. Traditional oligopoly theory suggests that it is easier to coordinate actions in a market with fewer players and, thus, post-merger coordination between C/D and the remaining two competitors in the market may reduce price, quality, or innovation competition.

On the other hand, most high-end data processing software customers state that the combination of C/D will likely create considerable efficiencies, particularly with respect to R&D. Most customers of A and B believe that combining the complementary R&D capabilities of C and D will enable the merged firm to produce higher-quality and a greater number of variations on its software and to innovate at a faster pace, thus making the merged firm a more formidable competitor against companies A and B.

The parties allege between $5.5 and $8.5 million in synergies and cost savings from consolidation of R&D efforts, customer service operations(6), and headquarters operations. One question is whether such efficiencies would be achieved by each company unilaterally in “the future likely without the merger.” Company C’s internal, pre-merger documents reveal that it already has accomplished some cost savings through streamlining customer service and headquarters operations, but that further cost savings are likely to be difficult unless there is a higher volume of sales over which to spread Company C’s fixed costs for these operations. Company D’s documents discussing the pros and cons of the proposed merger note various ways in which a merged C/D could combine the differing strengths of each company’s R&D efforts to create new options for users of high end data processing software programs. These facts, when combined, tend to suggest that further cost savings might be achieved through combining the customer service and headquarter operations of C and D, and that the proposed merger would likely enhance R&D efforts in the relevant market above the R&D efforts likely to take place in the hypothetical probable future without the merger.

Some market conditions further suggest that the proposed merger, although resulting in net fewer players, indeed will likely increase competition. Some software programs have unique programming qualities that make some of the products in the relevant market fairly heterogeneous. Orders for the software tend to be large and relatively infrequent. Thus, depending on other factors, coordinated interaction, tacit or express, may not be easy in the relevant market. Further, customers of C and D have stated that they have previously used other suppliers and would be willing to turn to A or B again if C/D attempted to raise (quality- adjusted) price or lower (price-adjusted) quality post-merger.

In conclusion, although the merger of companies C and D would increase concentration in an already highly concentrated market, this thumbnail sketch suggests that, overall, the transaction likely will generate substantial efficiencies that will result in a more competitive, or not less competitive, market post-merger.

Example #2 -- Industry in Transition. There are currently three competitors -- A, B, and C -- that produce and sell auxiliary power units that are necessary to provide maintenance on an older generation of a particular type of turbine engine. Companies B and C propose to merge, thus reducing the number of competitors from three to two.

Within the next three years, the older generation of engine will be superseded by the next- generation turbine engine of its type. This transition has been long anticipated, since the development of the new engine has been going on for almost ten years. The new engine will

require a next-generation auxiliary power unit; current units cannot generate sufficient power for the maintenance requirements of the new engines.

Because of the anticipated switch to new engines, sales of the older generation engines have declined substantially. Correspondingly, sales of auxiliary power units for the older generation engines have declined substantially. Company A has long been the market leader in sales of auxiliary power units and has become even more dominant in the last five years, gradually increasing its market share from 40% to 60%, as more and more customers have concluded that A is the only company in the relevant product market that is virtually certain to survive the transition to the next generation of auxiliary power unit. Thus, Company A has been able to continue its R&D plans for the next-generation auxiliary power unit, since its per unit costs of production have remained relatively stable. Company A’s R&D program has produced a prototype auxiliary power unit that appears to require only relatively modest changes to be successfully marketed to serve the new engine.

By contrast, Companies B and C have fallen behind, with their percentage of sales declining from roughly 30% each five years ago to roughly 20% each currently. As sales volumes have dropped, each company has found itself saddled with substantial excess capacity and dramatically rising costs per unit. Customers complain that Companies B and C have become less and less competitive on price with Company A. Customers also express concern that, although B and C each have their own R&D programs for a next-generation auxiliary power unit, neither program has yet produced any prototype for the customers to examine.

Entry into either the current or the next-generation auxiliary power unit market is unlikely to be timely, likely, or sufficient to deter or counteract any anticompetitive effect of concern. Given the substantial and continuing decline in sales for the current generation power unit, there would be little incentive for a potential entrant to make the required investment. For the next- generation power unit, entry also would not be timely, likely, or sufficient. It has taken Company A more than three years to develop its prototype for the next-generation auxiliary power unit, and it would likely take a new entrant even longer. Moreover, for a variety of reasons, current projections suggest that the market for sales of the next-generation auxiliary power unit will remain as small as the current market for the existing generation of power units. Thus, it appears questionable whether the proposed merger would generate sufficient sales opportunities for a new entrant to enter at minimum viable scale.

Companies B and C claim that the proposed merger would generate efficiencies that will enable them once again to compete successfully against Company A on price and that will enable them to produce a prototype of the next-generation unit on a much faster timetable. They propose to combine their operations in one manufacturing facility and thus achieve substantially reduced costs of unit production. The evidence on economies of scale in this and similar industries tends to support that such a combination in fact would significantly reduce unit production costs. In addition, an evaluation of the R&D programs of Companies B and C suggests that each has solved a different portion of the problems associated with the next- generation auxiliary power unit, so that a combination of their R&D efforts might well enable the merged firm to produce a prototype at a faster pace than either company would be able to achieve on its own.

So far, the facts overall appear to suggest that the proposed merger might not substantially lessen competition and might even increase competition. But the description so far does not include a thorough evaluation of all of the facts that could be relevant to a likelihood of post-merger coordinated interaction or unilateral anticompetitive effects. What if there was a history of collusion among the power unit producers? What if, based on such a history, customers reported some concern about the possibility of Company A and the merged firm adopting a customer allocation scheme for sales of the next-generation auxiliary power unit? Would the likelihood of cost savings and faster R&D then appear to be likely to increase competition between A and the merged firm, or would it appear simply to set the stage for a comfortable arrangement of tacit or express collusion between the two firms? What about the likelihood of unilateral anticompetitive effects? Should an agency consider a duopoly of relative equals preferable to a possible evolution of Company A as monopolist? Under the facts as outlined above, does it appear likely that, absent the proposed merger, Company A would become a monopolist or be able to produce anticompetitive effects unilaterally?

Suppose we changed some of the factual assumptions about the efficiencies and other evidence. What if neither Company B nor Company C had ever considered how to achieve cost savings unilaterally, and it appeared that unilateral cost savings of significant amounts would be practically feasible and reasonably achievable? What if the R&D programs of each of the merging parties were quite similar, and both companies’ internal documents projected the ability to produce a prototype within the next six months? What if it appeared that the market for next- generation auxiliary power units would be much larger than the auxiliary power unit market ever had been previously, implying that current excess capacity could be used cost effectively once the transition into the next-generation unit was accomplished?

Some facts could suggest the likelihood of anticompetitive effects post-merger, some could suggest the likelihood of procompetitive effects post-merger, and some could be argued in either direction. My point in raising all of these questions is just to remind us all how fact- specific merger analysis is and necessarily must be under current antitrust thinking. If the efficiencies proposal in the FTC staff report were adopted in some form, it would still be applied in the context of complex facts that do not always point in one direction. In that context, the results of applying the report’s analytical framework could vary substantially, depending on the overall assessment of all of the factors relevant to merger analysis.


In conclusion, let me say that we would welcome your thoughts and comments on the proposed approach to analyzing merger efficiencies that is set forth in the FTC staff report. The joint FTC/DOJ task force is meeting regularly to study whether to propose some revision to the efficiencies section of the current guidelines, and so comments and reactions to the staff report proposal or thoughts about other approaches to the merger analysis of efficiencies would be timely.

Thank you very much. I would be happy to take any questions.

(1)I am grateful to Debra Valentine and Gary Zanfagna (formerly in Policy Planning at the FTC) for assistance in preparing this speech and to George Cary, Bill Cohen, Melissa Heydenreich, John Hilke, Bob Leibenluft, and Howard Morse for helpful comments.

(2)Section 6 of the Federal Trade Commission Act of 1914, 38 Stat. 717 (current version at 15 U.S.C.A. § 46 (1988)).

(3)See Oliver Williamson, Economies as an Antitrust Defense: The Welfare Tradeoffs, 58 AM. ECON. REV. 18 (1968).

(4)See, e.g., United States v. Citizens & S. Nat’l Bank, 422 U.S. 86, 120 (1975); United States v. Marine Bancorporation, 418 U.S. 602, 631 (1974).

(5)Timing is an important issue, and one that will have to be addressed should guidelines changes be made. Under our proposal, the agency would employ a sufficiently flexible time frame to capture adequately the dynamic effect of efficiencies that likely would contribute to a more -- or no less -- competitive market post-merger.

(6)As was earlier discussed in connection with possible claims of cost savings from no longer providing customers with free samples, cost savings in customer service operations might or might not reflect likely procompetitive effects. One relevant question would be whether the merged firm planned to provide the same level of customer service through more efficient operations or planned to cut back on the level of customer service that was offered.