The Role of Efficiency Considerations
in the Federal Trade Commission’s Antitrust Analysis
Morgan, Lewis & Bockius LLP
Presented at the Federal Trade Commission’s
Hearings on Global and Innovation-Based Competition
November 14, 1995
Efficiency considerations have assumed an important role in antitrust analysis over the past 15 years. Although traditional rule of reason analysis has long entailed a balancing of anticompetitive effects of a course of conduct or transaction against the procompetitive benefits associated with the conduct or transaction, it was only the Chicago revolution in the late 1970s that placed efficiencies at the center of antitrust analysis. Yet even today, one must search far and wide to find a case in which efficiencies, described and analyzed as such, trumped a strong anticompetitive story. The problem has been, as Chairman Pitofsky has observed, one of proof: claims of efficiencies are “easy to assert and sometimes difficult to disprove.”(1) To this observation one should add the corollary that valid efficiency claims can be extraordinarily difficult to prove.
How, then, should antitrust enforcers deal with the question of efficiencies, that often-neglected significant other of antitrust analysis? How can efficiencies have such a central role in the analysis if specious, but plausible, efficiency claims are easy to assert and difficult to disprove while many valid claims are exceedingly difficult to prove, if they are at all susceptible to proof. In the area of merger enforcement, the antitrust enforcement agencies have found the right balance by adjusting the core analysis of competitive effects to account indirectly for the enhanced likelihood that certain transactions are likely to bring about efficiencies. It is no accident that we seldom see a challenge of a defense industry merger, hospital merger, or software merger where four or more surviving competitors remain. In industries in which scale economies are significant, the agencies have routinely, and quite correctly, adjusted the numerical presumptions of the Horizontal Merger Guidelines to accommodate the scale of enterprise needed to attain efficient (though not optimal) scale, whether that scale is based on research and development efficiencies, production efficiencies, or even (in the health care area) administrative and overhead efficiencies.
This approach has allowed the agencies to circumvent in most cases the difficult problems of proof and yet accommodate in some fashion the need to credit the efficiency side of the ledger. At the same time, the agencies have been increasingly receptive to, though still skeptical of, claims that specific transactions are necessary to achieve more efficient operations. This is a sound approach so long as the analytical tools available to antitrust analysts remain too crude to separate the wheat from the chaff of efficiency claims. Efficiency claims should be credited where they can be substantiated, and the agencies should be willing to hear out proponents of efficiencies. If policy adjustments are called for, they are modest indeed.
On the nonmerger side, the story is more complex. Efficiency considerations are explicitly made a significant component of rule of reason analysis. The hornbook explanation of the rule of reason is that it calls for balancing of pro- and anticompetitive effects. The enforcement agencies adopted the right approach in the new Antitrust Guidelines for the Licensing of Intellectual Property, where the analytical framework for evaluating restraints on competition calls for evaluating “whether the restraint is likely to have anticompetitive effects and, if so, whether the restraint is reasonably necessary to achieve procompetitive benefits that outweigh those anticompetitive effects.”(2) Yet the same problems of proof that plague efficiency analysis in the merger area make that balancing act an illusion. Courts, and to lesser extent, the agencies have tended to find conduct either pro- or anticompetitive. Seldom has a court grappled with evidence that shows that the same conduct produces meaningful pro- as well as anticompetitive effects.
The sole area in which recent antitrust policy with respect to efficiencies can be called a failure is, I am sad to say, a creature of this agency. Under the Commission’s so-called Mass. Board analysis,(3) the agency places the onus on the respondent to come forward and prove that its conduct will produce efficiencies that are both “plausible” and “valid” in order to escape antitrust condemnation whenever the Commission concludes that a practice is “inherently suspect.” Given the enormous difficulties that can be associated with proving valid efficiencies, the effect of the approach is to condemn conduct that may or may not be efficient but has not been shown to have any anticompetitive effects solely because the respondent has failed to overcome the complex problems of proving efficiencies. This is a fundamentally wrong approach to antitrust, and the Commission should use the occasion of these important hearings to renounce the use of this anticompetitive doctrine.
I. Efficiency Considerations in Merger Analysis
The Supreme Court’s 1977 Sylvania decision(4) is a landmark opinion not only for the changes it brought about in the analysis of vertical restraints but for the way in which it brought efficiency considerations into the forefront of antitrust analysis. For the first time, a court looked at the efficiency-enhancing capacity of a business practice through the lens of economic analysis and declared that efficiencies were central to an understanding of the competitive significance of the practice. Since then, the Supreme Court has reaffirmed the central role of efficiencies in analyzing competitive restraints in a spectrum of cases ranging from BMI(5) to Northwest Stationers.(6) Yet today, 18 years after Sylvania, there has been no merger case in which efficiencies have played such a similarly central role.
To date, the Supreme Court has not had the opportunity to reconsider its Philadelphia National Bank decision, where it held that a merger that is deemed likely to lessen competition “is not saved because, on some ultimate reckoning of social and economic debits and credits, it may be deemed beneficial.”(7) This is unfortunate, because that decision represents a different era in antitrust, an era in which formalistic rules rather than economic reasoning ruled the day and efficiencies were often thought harmful. It boggles the mind today to observe that a mere 20 years ago, this Commission found it necessary to declare that “improved efficiencies and price reductions are certainly no reason to condemn a merger not otherwise shown to be anti competitive.”(8) Worse, decisions both by the Commission itself and by its administrative law judges suggested that efficiencies tend to support a holding of illegality and that their absence tends to favor a holding of legality.(9)
We have come a long way from Philadelphia National Bank and the era in antitrust that produced it. Today the Merger Guidelines make it clear that it is possible, at least in theory, for efficiencies to outweigh competitive concerns and thus save an otherwise anticompetitive merger from condemnation.(10) Yet the day has not yet come when either a court or an agency would declare that a particular merger that is otherwise anticompetitive may go forward because it is likely to generate efficiencies that will outweigh the effects that are expected to flow from a substantial lessening of competition in the relevant market.
In any event, courts have not found it necessary to balance potential anticompetitive effects of mergers against efficiencies. Courts have tended to align efficiencies with their determination of the likely competitive effects of challenged transactions. Courts have tended to reject efficiency claims on evidentiary grounds in cases in which they found mergers to be anticompetitive and to credit claimed efficiencies when sustaining transactions on competitive grounds.(11) Last month’s decision in United States v. Mercy Health Services(12) is in a small minority in deviating from this pattern, rejecting both the government’s claim of anticompetitive effects and the merging parties’ efficiency claims.
The antitrust enforcement agencies’ tendency to reject efficiency claims in cases in which they have concluded that a merger is likely to reduce competition is not a reflection of an inherent hostility to the efficiency defense. The defense has now been enshrined in the Merger Guidelines for over a decade. In this area the agencies were a step ahead of the courts in recognizing an efficiency defense. The outcome does reflect, however, tremendous skepticism among enforcers to efficiency claims, which can be traced at least in part to the large number of specious efficiency claims that an enforcement lawyer hears in the course of a career. As a consequence, even strong efficiency claims are often met with skepticism.
The problem of proof lies in demonstrating that claimed efficiencies cannot be attained through means that are significantly less restrictive of competition, as the Merger Guidelines require. This is a sound requirement as a matter of policy — an anticompetitive transaction should not be excused on the basis of efficiencies that could be accomplished without harm to competition — but it comes at the cost of enhancing the burden for parties that have legitimate efficiency claims. Proof of efficiencies becomes a matter of detailed and costly econometric studies that seek to quantify both the efficiencies that can be attained through a merger and those that can be captured in its absence and to isolate those that can be captured only through the transaction.
This proof, of course, gets you only half way to proving your case. The Guidelines require cost savings from an otherwise anticompetitive transaction to be passed on to consumers in the form of lower prices, and this entails further elaboration of the econometric work to compare the prices that are likely to be charged in the but-for world with those that prevail before the merger. This is a tall order, yet it is a matter of necessity to assert a full-fledged efficiency defense. Yet very often the data necessary to proffer this proof simply will not exist. In other cases, the data might exist but the cost and time required to complete an efficiencies study may be unacceptable.
Why then do practitioners insist on doing the impossible by proffering efficiency defenses in cases in which the agencies’ exacting standards cannot be met, or at least cannot be met economically? Because it is important for enforcers to understand the business reasons that underlie transactions. Even if it is impossible for an acquirer to demonstrate in a timely manner that $25 million in annual savings in research and development will be passed on to consumers in lower prices, showing the investigating agency that these savings will be attained will dispel the mistaken notion that the $10 million price premium paid by the acquirer reflects some kind of monopoly or oligopoly premium.
The agencies understand the problems of proof quite well and have adjusted enforcement policies to enable merging parties to capture efficiencies in markets in which there is a strong case to be made for consolidation based on scale economies. Indeed, Chairman Pitofsky has been a leading proponent of adjusting antitrust policy to accommodate the special conditions of distressed industries.(13) As the defense industrial base contracts, it makes sense to have two plants operate at something resembling efficient scale than to have three struggling plants engaged in a survival-of-the-fittest competition. As hospitals face an ongoing reduction in patient days, consolidation of wards or administrative functions through mergers will often make more sense than a struggle to survive at low utilization rates. And in R&D intensive markets, a reduction in the number of independent research efforts may be beneficial not only because it will lead to cost savings but also because it may hasten, rather than impede, the likelihood of successful innovation. That the federal antitrust agencies recognize these realities is evident from the way in which they have enforced the merger laws in the defense, health care, and software industries.
The challenge for the enforcement agencies is to lend a receptive ear to efficiency claims and adjust the competitive effects analysis in other industries in which efficient operations may require a market structure with concentrations, measured by the Herfindahl-Hirschman Index, in excess of the levels that ordinarily raise red flags. If R&D expenditures in an industry must be spread over a large number of purchases for a company to be an efficient competitor, the agencies should recognize this fact and adjust their presumptions, irrespective of the industry involved. If decline in demand for a product has produced factories that are operating at 50 percent of capacity, the agencies should recognize that the economy is better off with fewer plants operating at a more efficient scale than with more plants and large unused capacity, regardless of the industry involved. The agencies do this already as a general matter, although it is probably a fair statement that they are more comfortable in accommodating these concerns in some industries than in others. The approach currently in use ought to be used broadly, wherever it makes economic sense.
One cost that is inherent in the current approach, and the approach endorsed here, is a loss of transparency. The approach, of course, is to apply the Merger Guidelines in a nonliteral fashion, which means that the Guidelines lose some of their value as a predictive tool for the private sector. It takes an inside-the-beltway guru to tell you that 1,800 and 1,800 are not necessarily the same, or that 2,000 in market A may be worse than 3,000 in market B. This is true to a large extent in any event, but the recognition of distressed industries or R&D-intensive industries may exacerbate that. As the agencies gain more experience with these issues both in the context of declining or failing industries and in the context of R&D-intensive industries, a time will come to revise the Guidelines to reflect that experience.
II. Efficiency Considerations in FTC Nonmerger Cases
In evaluating the treatment of efficiencies in horizontal restraint cases by the antitrust agencies and the courts, anyone who practices before the FTC is immediately struck by the fact that there are two antitrust worlds out there: the FTC and the rest of the world. The judiciary and the Justice Department follow the approach of evaluating the anticompetitive effects of a course of conduct, including an evaluation of market power, and addressing the issue of efficiency justifications only if that first stage of the analysis indicates that the conduct is likely to harm competition. This is the approach articulated in the new Antitrust Guidelines for the Licensing of Intellectual Property, of which the FTC is a co-sponsor.(14) Given the difficulties inherent in establishing the existence of efficiencies in the nonmerger context, which will be discussed shortly, this is the only sensible approach.
The Commission, unfortunately, adopted in 1988 a radically different approach to horizontal restraints, the so-called Mass. Board approach. Under this approach, the Commission asks whether an agreement is “inherently suspect,” that is whether, absent an efficiency justification, it would tend to restrict output. If an agreement is inherently suspect, it then asks whether it has a plausible efficiency justification. Absent a plausible justification, the suspect agreement is condemned without an analysis of its competitive effects. And, even if an agreement has a plausible efficiency justification, the Commission will nevertheless condemn it without conducting a competitive analysis if it deems the justification invalid. Only if the agreement is not inherently suspect to begin with or has a valid efficiency justification will the Commission engage in a more extensive rule of reason inquiry.
That the creators of Mass. Board were well-intentioned is not in doubt. As they articulated the rationale for the decision in contemporaneous discussions, they conceived the approach as a way to capture the near-per se cases, the types of cases that the Supreme Court in Indiana Federation of Dentists analyzed summarily, without a detailed review of market structure.(15) Yet in implementation, Mass. Board has become a substitute for analysis. The Commission has never given content to the term “inherently suspect,” which triggers the Mass. Board analysis, and unfortunately that trigger has been pulled quite often by the staff, with the best of intentions, because of the convenience that it offers. It is much easier to write a memo that explains why a business practice should be deemed inherently suspect than to conduct an investigation of the competitive effects of the practice. In the many years in which I served in an evaluation function at the Bureau of Competition, I seldom reviewed a staff recommendation that analyzed a horizontal restraint under the rule of reason. Mass. Board was the norm, not the exception.
This implementation record tells us that at the Federal Trade Commission efficiencies are not a defense but a part of the respondent’s affirmative burden of proof; unless the respondent can prove that a practice deemed “inherently suspect” is supported by a valid efficiency, the practice is condemned even though the government has not put forward one iota of proof that the practice harms competition. This is truly extraordinary. And it has real-life effects. One need only leaf through the consent orders that the Commission has secured since 1988 in conduct cases to realize that much of the conduct that was the subject of the accompanying complaints had no potential for causing injury to consumers. Each complaint that lacks an allegation of market power — and very often complaints do not even allege that the respondents possessed a significant market share, let alone market power — tells a story of antitrust condemnation of competitors that lacked the ability to harm competition.
Why is this so important? First, the burden-shifting embodied in the Mass. Board approach is contrary to the basic notion of Anglo-American jurisprudence that the plaintiff must bear the burden of proof. Second, the approach is virtually tailor-made, albeit as a result of a design flaw rather than intention, to yield incorrect outcomes. And this effect is compounded by the tremendous uncertainty that the approach engenders. Consider a transaction in which two competitors collaborate to develop a new product and market the product jointly. They agree not to market the product in competition with the collaboration. There is little doubt that this agreement can be used to effectuate an anticompetitive scheme, but there should be equally little doubt that the agreement will have a procompetitive effect in most cases. This is because the parties will often have a legitimate need to protect the joint undertaking from free-riding by their co-venturer. A party is more likely to contribute its technology to the collaboration if it can be protected from having its co-venturer use its own technology against it than if it is bereft of that protection.
Is this a case that the FTC will evaluate under Mass. Board? One hopes not and one thinks not, but one cannot say with confidence that the case will be evaluated by the FTC under the rule of reason. The mode of analysis that the Commission will use hinges on whether the restraint will be deemed inherently suspect, but the term “inherently suspect” is inherently elastic. And if the burden is placed on the respondents to justify the restriction, what must they show to validate the efficiency claim? The efficiency claim just outlined is certainly a plausible one, but Mass. Board demands strict proof that the justification is “valid” in the circumstances of each particular case. And it just may be that one party’s insistence on the restriction at issue is simply the product of one corporate executive’s hunch that the co-venturer may engage in opportunistic behavior in the absence of the restriction. Is that a valid efficiency? It ought to be, but one does worry that the same approach that finds business practices “inherently suspect” will be less than wholly receptive to efficiency justifications based on hunches.
In the business world, as in the government, decisions are often made on the basis of incomplete information. A business executive may decide that a collaboration with a competitor must limit the opportunities for free-riding without having complete information of whether the other party may try to take a free ride or even has the capacity to do so. Yet this should not matter a whit in the analysis. All that should matter for purposes of the efficiencies analysis is that a procompetitive transaction likely would not have taken place in the absence of the limitation. Yet proving even that much is often a difficult proposition, as the reasons for the inclusion of a particular restraint in an agreement may reside only in a corporate executive’s mind.
To be sure, these problems of proof, these uncertainties, will be present no matter what the standard is whenever the issue of efficiencies is at issue. But Mass. Board puts the issue front and center before there is any evidence that the restriction at issue has had any competitive effects. That is the fundamental problem with Mass. Board. The approach forces respondents to establish the difficult proof of innocence (or efficiency) before the FTC is called upon to show that there is reason to ascribe guilt (adverse competitive effects).
These are not simply the views of a private practitioner. It is a fair reading of the DOJ/FTC intellectual property guidelines that these concerns are shared by the Justice Department’s Antitrust Division. One need not have been privy to the negotiations that must have transpired between the agencies to deduce that the DOJ rejects the Mass. Board approach. The Guidelines, which by their terms “state the antitrust enforcement policy of the U.S. Department of Justice and the Federal Trade Commission,” explicitly inform the reader that Mass. Board represents “the Federal Trade Commission’s approach.”(16) This is truly extraordinary for joint guidelines. It is the single instance in which one of the antitrust agencies has publicly rejected the other’s analytical approach within the context of such guidelines. And it is most fitting that the DOJ would do so within the framework of an attempt to expand the reach of Mass. Board to the intellectual property area.
If the Commission takes no other action as a result of these hearings, renouncing the Mass. Board standard should be the one action that it does take. This Commission has recently undertaken some bold actions to reduce burdens on the private sector in greatly limiting the use of the prior approval remedy in merger cases and in terminating outdated orders. It has also taken the commendable step of evaluating the efficacy of merger remedies that it has used. These steps have earned the Commission much-deserved goodwill and respect in the antitrust community. Revisiting Mass. Board would add further credit to the Commission.
(1) Partner, Morgan, Lewis & Bockius LLP, Washington, D.C.
(1) Robert Pitofsky, Proposals for Revised United States Merger Enforcement in a Global Economy, 81 Georgetown L.J. 195, 210 (1992).
(2) 6 Trade Reg. Rep. (CCH) ¶ 13,132, § 3.4 (April 6, 1995).
(3) See Massachusetts Board of Registration in Optometry, 110 F.T.C. 549 (1988).
(4) Continental T.V., Inc. v. GTE Sylvania Inc., 433 U.S. 36 (1977).
(5) Broadcast Music, Inc. v. Columbia Broadcasting Systems, Inc., 441 U.S. 1 (1979).
(6) Northwest Wholesale Stationers, Inc. v. Pacific Stationery & Printing Co., 472 U.S. 284 (1985).
(7) United States v. Philadelphia National Bank, 374 U.S. 321 (1963).
(8) Beatrice Foods Co., 86 F.T.C. 1, 66 (1975), aff’d, 540 F.2d 303 (7th Cir. 1976).
(9) Wesley J. Liebler, Antitrust Law and the New Federal Trade Commission, 12 Sw. U. L. Rev. 166, 225 (1981).
(10) Horizontal Merger Guidelines, 4 Trade Reg. Rep. (CCH) ¶ 13,104, § 4 (April 2, 1992).
(11) See Joseph Kattan, Efficiencies and Merger Analysis, 62 Antitrust L.J. 513, 519-20 (1994).
(12) No. C94-1023 (N.D. Iowa Oct. 27, 1995).
(13) See Robert Pitofsky, supra note 1.
(14) See Antitrust Guidelines for the Licensing of Intellectual Property, 6 Trade Reg. Rep. (CCH) ¶ 13,132, § 3.4 (April 6, 1995)
(15) FTC. v. Indiana Federation of Dentists, 476 U.S. 447 (1986).
(16) Intellectual Property Guidelines, §§ 1.0, 3.4 n.27 (emphasis added).