Charles River Associates Incorporated, Conference on Economists' Perspective on Antitrust Today, The Boston Mariott Hotel
Note: This speech has been revised and is published in the Antitrust Law Journal, vol. 65, Winter 1997, pp. 353-374.
My subject today is a decision that has not yet been written by the Supreme Court. It is a decision that must be written sooner or later, though, to remove a shadow that has fallen over the treatment of entry in merger analysis under Clayton Act §7.
The problem the Supreme Court must solve is created by two appellate opinions handed down in 1990: one by the D.C. Circuit in Baker Hughes and the other by the Ninth Circuit in Syufy.(1) These decisions encourage an analytical confusion that potentially creates loopholes through which anticompetitive mergers may pass. But I am getting ahead of my story. I want to begin by describing an earlier time when the circuit courts took the lead in correcting antitrust’s neglect of supply-side forces before I explain how the lower courts may have more recently begun to go off track. And, especially when I discuss the opinions of eminent judges, the people I work for insist that I remind you that my remarks do not necessarily reflect the views of the Commission or any individual Commissioner.
The Supply-Side Before Chicago
Antitrust has long understood that a firm’s attempt to exercise market power could be defeated by new supply entering the market. Entry, or just the potential for entry, might prevent even so-called monopolists from actually exercising market power. Thus, the Supreme Court’s Columbia Steel decision, handed down a half century ago, rejected the government’s challenge to the Kaiser-Columbia Steel acquisition in part on the ground that supply substitution resulted in a broad product market.(2) Even in a Supreme Court merger decision that did not make the Chicago school’s hit parade, such as Brown Shoe, the competitive importance of supply substitutability was recognized — albeit as dictum relegated to a footnote.(3) And the product market definition in the Supreme Court’s 1966 Grinnell decision, a monopolization case, arguably accounted for supply substitutability, although the Court did not acknowledge this interpretation.(4)
Until the mid-1970s, however, these important insights about supply substitution (or production flexibility, as it is also called) were neglected by mainstream antitrust jurisprudence. In the postwar, pre-Chicago era, antitrust product market definition was based almost exclusively on demand substitutability, consistent with the doctrinal formulation of the Supreme Court’s Cellophane case that product markets are collections of goods with “reasonable interchageability” in demand.(5) Judicial focus on demand substitutability was reinforced by the Court’s 1964 Rome Cable decision, which placed insulated copper conductor and insulated aluminum conductor in separate markets, despite the strong dissent by Justice Potter Stewart highlighting the extensive supply substitutability between the two.(6) Given the Court’s emphasis on demand substitution, the federal circuits were understandably reluctant during the 1960s to accept supply substitution as a basis for product market definition, even when confronted by reasonable economic arguments for doing otherwise.(7)
The Supply-Side Revolution in the Circuit Courts
During the 1970s, lower courts took the lead in moving antitrust to recognize the role of new supply in constraining the exercise of market power. Indeed, judicial adoption of the economically-oriented Chicago school approach — the distinctive feature of the contemporary antitrust landscape — arguably began in 1975 in the appellate courts on the issue of supply substitution. The leading modern decisions recognizing the role of supply substitution in market definition were issued in separate circuits less than one month apart: the 10th Circuit’s Telex opinion and the 9th Circuit’s opinion in Twin City.(8) This happened two years before the Supreme Court handed down GTE Sylvania and Brunswick, and four years before BMI, to name some of the other early Chicago school landmarks.(9)
Neither Telex nor Twin City involved a merger, but monopolization allegations called for market definition in each. Telex found a market for all computer peripherals rather than only IBM plug-compatible ones. This made no sense from the demand side: a consumer could not substitute a peripheral device that was incompatible with its existing large and expensive computer. But a supplier could add compatibility, probably without much difficulty.(10) Similarly, Twin City questioned a lower court’s market for “concession services for major league baseball” that did not consider how concession operators at other kinds of facilities could offer the same service.(11) Again, from the demand side this does not make sense: the fan at the ball park could not substitute a hot dog from the vendor at the zoo. But the operator of the zoo’s food concessions could surely figure out how to serve the ball park without much trouble. After these decisions, supply substitution was rapidly incorporated into product market definition by other circuits and the FTC, and the new approach was endorsed in commentary.(12)
This supply-side Chicago revolution was accomplished by the circuit courts with little instruction or guidance from the Supreme Court. To be sure, General Dynamics had come down the year before.(13) But that case did not turn on supply substitutability. General Dynamics instead holds that the presumption that concentration lessens competition, set forth in Philadelphia National Bank, can be rebutted by evidence that a merging firm’s historic market share overstates its future competitive significance.(14) This sounds like a Chicago school decision more in retrospect than it likely appeared at the time,(15) and the appellate panels did not rely on it to justify incorporating supply substitution in market definition. The circuit courts instead reached back to Columbia Steel, the Brown Shoe footnote, and Grinnell for authority.(16) But they were really acting on their own in recognizing that a monopolist or cartel cannot successfully raise price above competitive levels if enough other firms could readily alter their production processes to make a competing product. They took that first, most difficult step away from Egypt toward the Promised Land.
Uncommitted Entry in the Merger Guidelines
The 1982 Merger Guidelines, drafted a few years after Telex and Twin City, took a different approach to incorporating supply responses in merger analysis, and that approach has been maintained and more fully articulated by the 1992 Guidelines.(17) In the Guidelines, supply substitution is not considered in defining markets. Markets are defined instead based solely on demand-side substitutability considerations, much as the Supreme Court suggested in the Cellophane case. But unlike the courts in the pre- Chicago era, the Merger Guidelines consider the possibility that new supply would defeat the exercise of market power.
To discuss the role new supply plays, it is useful to adopt the terminology of the 1992 Merger Guidelines, which distinguish between “uncommitted” and “committed” entry. The concept of uncommitted entry generalizes the idea of supply substitution. Uncommitted entry is hit-and-run. Uncommitted entrants are firms that can enter quickly, and with little sunk expenditure. They can take advantage of any short-run profit opportunities that anticompetitive behavior by incumbent firms might offer, and get out rapidly and cheaply if those opportunities disappear. Committed entrants, in contrast, are in for the long haul. Once they enter, they expect to stay, because to abandon the market would mean walking away from a substantial sunk investment. Because they are in for the long haul, they must consider what competition will look like after they enter in deciding whether it is profitable for them to enter in the first place.
The 1992 Guidelines recognize several forms of uncommitted entry. The simplest is production substitution, for example using a canning plant to pack fruit instead of pudding, using a retail space to sell sweatshirts instead of shoes, or using a machine shop and engineering team to make heavy duty drills instead of milling machines. The Guidelines also recognize that a brand name might also be applied quickly to a new use without significant additional sunk expenditures — perhaps the way Minute Maid, known for orange juice, added carbonation to create a soft drink product.(18) And, if de novo entry can be accomplished quickly and at little sunk cost, that too would constitute uncommitted entry.
Under the 1982 and 1992 Guidelines, uncommitted entry is important in determining the identity of market participants and market shares, but not in defining the market itself. The scope of the relevant product market is determined entirely by demand substitution. In contrast, the Telex and Twin City approach to recognizing supply substitution expands the market.
While the appellate panels had the right insight — that supply substitution can constrain the exercise of market power — and adopted a simple approach to incorporate it into antitrust analysis, the Guidelines found a better means of implementation. To see the difference, consider a hypothetical merger of firms selling aluminum cable; assume that aluminum cable and copper cable are not demand substitutes; and assume that manufacturers of copper cable can quickly switch their cable capacity to produce the aluminum product with little or no sunk expenditure. In thinking about whether the price of aluminum cable would rise, it is necessary to take into account the competitive constraint imposed by the copper cable producers. The circuit courts would do so by defining an all cable market, and calculating market shares that ignore the difference between copper and aluminum cable. Even if aluminum cable production is highly concentrated today, the all cable market could be unconcentrated, in which case the courts would infer that the aluminum cable producers could not exercise market power.
That may often be the right answer, but not always. The market definition approach to accounting for supply substitution is biased toward underestimating the potential for exercise of market power in aluminum cable because it implicitly assumes that in the event aluminum cable prices were to rise, all copper cable production assets would be available to shift supply into the aluminum cable market. This assumption may not be realistic because the opportunity cost of diverting cable from copper to aluminum may be high. To be sure, if copper cable capacity is not being utilized, and can inexpensively be pressed into service to make aluminum cable, supply substitution may come at no opportunity cost. But that is not the only possibility. It may be costly to divert copper cable production tied up in long term contracts; even absent contracts, a copper cable supplier may be reluctant to undermine its long term customer relationships by cutting off copper cable buyers to make aluminum cable sales; and copper cable producers may find that the price of copper cable rises as production shifts to aluminum cable, raising margins in the copper cable market and increasing the opportunity cost of shifting production to service aluminum cable buyers. When diversion is costly, it can be misleading to act as though all copper cable production assets would be available to compete away a rise in the price of aluminum cable by computing market shares for an all cable market.
The Merger Guidelines avoid the all-or-none problem created by the circuit court approach to accounting for supply substitution by assigning market shares to uncommitted entrants based on divertible capacity(19) — in the example, the capacity the copper cable producer would find profitable to shift into aluminum cable production in the event of a small aluminum cable price rise. The resulting market shares will avoid making concentration look low when little copper cable capacity would readily shift into the production of aluminum cable. They will thus avoid underestimating the competitive problem from a merger of two aluminum cable producers under such circumstances.
The Guidelines’ approach also avoids overestimating the potential for entry to deter or counteract the competitive problem from merger, unlike the circuit court approach. The problem arises if an entrant has a new technology for making copper cable that — unlike the existing production processes — would not permit the firm to produce aluminum cable. The prospect of that entry would may well be insufficient to solve a competitive problem resulting from a merger of aluminum cable producers, but the circuit court approach of defining an all cable market would not naturally rule that suggestion out. Moreover, it is hard to delineate price discrimination markets based on the idea of targeted buyers, as the Merger Guidelines permit, unless markets are defined based on demand-side substitution.
Despite these problems, it is hard to fault the appellate panels for choosing in 1975 to use market definition as the vehicle for incorporating supply substitution into antitrust analysis. The Telex and Twin City approach may have been easier and more natural than the Guidelines’ approach. After all, the Clayton Act seems to call explicitly for the definition of markets, while it says nothing about entry. And in many cases, there may be little practical difference between the results achieved through the two methodologies. In short, the lower courts led antitrust in the right direction in 1975, even though they employed a less than perfect compass.
Entry Analysis in the Courts Before Baker Hughes and Syufy
Around the mid-1980s, after addressing the problem of incorporating supply substitution into market definition, the lower courts and the enforcement agencies turned their attention to the role of entry in merger analysis. When they examined entry, the courts continued on the right track for a while, but they eventually began to take a wrong turn, potentially creating a loophole in merger enforcement. It is tempting to compare the lower courts to the Israelites, who followed a divine command to leave Egypt only to lose their spiritual direction while wandering in the desert by turning to worship a golden calf.
To explain the problem, we must begin in Egypt, or its antitrust counterpart: the Supreme Court’s 1963 decision in Philadelphia National Bank.(20) This decision establishes that Clayton Act §7 confers a presumption of anticompetitive effect on an acquisition that increases concentration in a concentrated market. The Court confirmed that conclusion in 1974, in General Dynamics.(21) Following these decisions, lawyers considering the role of entry in enforcing §7 — a topic which became interesting after the lower courts commenced their supply-side revolution in market definition — naturally asked: Does easy entry merely dilute or weaken the inference of anticompetitive effect? Or does easy entry trump everything else, so that once it is shown, the plaintiff must lose regardless of any other evidence?
Around the mid-1980s, the courts decided to treat easy entry as a trump.(22) That was the doctrinal implication of the Second Circuit’s 1984 Waste Management decision.(23) The district court had blocked this merger involving waste haulers, after concluding that easy entry merely weakens the inference of anticompetitive effect arising from concentration. The lower court noted that "there is no persuasive authority" for allowing low entry barriers and potential competition "to overcome a strong prima facie showing of concentration in the existing competitive structure."(24) The Second Circuit rejected this legal proposition and reversed the district court. It held that a finding of easy entry into trash collection trumps the other evidence relied upon below.
A year later, in the Calmar case, a district court reached a similar result: it refused to be concerned about a merger in pump sprayers once it found that any firm in the injection molding business could easily make them.(25) Meanwhile, the FTC decided Echlin, in which it dismissed a complaint despite high concentration in the carburetor kit product market on the view that any new entrant could set up shop in a backyard garage.(26) Legally, all three cases stand for the proposition that easy entry trumps high concentration — that ease of entry rebuts a prima facie case based on concentration data.(27)
In all three cases, the trump was applied to entry envisioned (rightly or wrongly) as uncommitted and unlimited — low sunk cost, rapid, hit and run entry by hordes of potential new competitors.(28) It was not necessary to attempt to analyze this uncommitted entry through market definition because, the courts found, entry was practically unlimited. Instead, the courts effectively observed that so many potential hit and run entrants were available as to make the market appear unconcentrated.
As of the mid-1980s, therefore, the courts had worked out an approach to dealing with uncommitted entry, whether limited or unlimited. They treated limited uncommitted entry possibilities — as when only certain specified supply substituters are poised to enter, each with a limited amount of divertible capacity — by expanding the product market definition. This was a plausible though imperfect approach to recognizing the competitive significance of supply substitution. And they recognized that unlimited uncommitted entry possibilities mean that mergers are unlikely to harm competition because markets are effectively unconcentrated. Thus, they treated unlimited uncommitted entry possibilities as a trump.
The mid-1980s entry cases technically left undecided the treatment of committed entry, which requires significant sunk investments and is in for the long haul. But the opinions never articulated a difference between committed and uncommitted entry, largely because they were not asked to do so.(29) By not noting such a distinction, these cases may have inadvertently led the circuit courts astray in Baker Hughes and Syufy. In both, the courts wrote as though they were explaining how the legal doctrine devised to handle unlimited uncommitted entry — that ease of entry is a trump — should applied to factual settings in which entry could be committed as well as uncommitted. The Baker Hughes and Syufy panels may have been correct in their application of the doctrine on the facts as they found them, for both appear to have concluded that the entry possibilities were uncommitted, contrary to what the government argued. But their analysis seemed to sweep in committed entry as well, thereby creating an analytical confusion that threatens to take entry analysis down the wrong path. To see how this happened and why it creates a problem, it is useful to begin by examining the economic analysis of committed entry in the 1992 Merger Guidelines.
Entry Analysis in the 1992 Merger Guidelines
On the surface, the 1992 Merger Guidelines might appear to take the same route I am suggesting the courts were wrong to follow. After all, they effectively declare ease of entry a trump, even when entry is committed. There is a subtle but critical difference, however, between the Guidelines approach, which I think is correct, and the more troublesome approach I will argue some courts may be suggesting.
The Guidelines say that committed entry is a trump — as long as the entry would be timely, likely, and sufficient.(30) A great deal of economic analysis is embedded into that qualification, and I will touch on some of it momentarily.(31) But I want to highlight first that from a legal point of view, the Guidelines should be read as insisting — quite properly — that entry should not be thought of as some abstract property of a market. In conducting merger analysis, it is not right to conceive of entry barriers as either high or low, and to set the courts off to determine which. Rather, the Guidelines formulation — timely, likely, and sufficient — emphasizes something simple and critical: committed entry is relevant to merger analysis only insofar as it will deter or counteract the competitive problem the merger would otherwise create. The language of Clayton Act §7, which makes a merger illegal only if its effect “may be substantially to lessen competition, or to tend to create a monopoly,” demands no less.
In a sense, the sufficiency element sums up the entire Guidelines analysis: what we care about is whether entry will be sufficient to cure a merger's competitive problem. This element is stated separately to highlight the possibility that even rapid and profitable entry might not be sufficient. It might be limited in “magnitude, character and scope,” to use another Guidelines phrase,(32) even if it is not delayed in time.(33)
Entry likelihood may have created some confusion, so I will step back from the case law evolution to spend a few moments discussing it. Entry likelihood analysis asks whether an entry plan would be profitable to carry out.(34) This is an important question because we care about entry only if it would cure the competitive problem from merger. An entry plan that would not be profitable for a prospective competitor to execute in the post-merger environment will not help deter or counteract anticompetitive conduct, even if it is technically feasible. In shorthand slogan, the Guidelines pick “would” over “could.” They care about whether a firm would enter, not merely whether it could enter.
The 1992 Guidelines articulated a then-novel conceptual framework for thinking about entry likelihood. The underlying principle is simple: committed entry is considered likely if it would be profitable in the post-merger environment. But recall that committed entry is in for the long haul. If the committed entry solves the competitive problem — that is why we are evaluating its profitability, after all — the post-merger price will quickly return to the pre-merger level or fall below it. Thus, the prospective committed entrant must determine whether its entry plan would be profitable assuming that it would receive no more than the pre-merger price.
This observation immediately raises questions, whose answers turn out to be the heart of the likelihood analysis. If entry at pre-merger prices would be profitable after the transaction, wouldn’t it have been profitable before? And wouldn’t the firm have entered already? In short, does the “likelihood” analysis guarantee that we will never find a likely entrant, because all likely entrants would already be incumbents?
The answer to the first two questions is “not necessarily,” and the answer to the last is “no.” Mergers change industry structure, and thus can change incumbent behavior and entrant incentives. The market after the merger is not the same as the market before it. The exercise of market power — what we fear will happen after the merger when we examine entry likelihood — leads to higher prices and lower industry output, and thereby creates additional potential sales for an entrant beyond what had been available before the merger. The result is to make entry more attractive than it had previously been; as the Guidelines put it, a merger can create an additional "sales opportunity" for an entrant.(35) Entry therefore may be profitable after the merger even if it had not been profitable before.
But is the additional sales opportunity resulting from a potentially anticompetitive merger large enough to make entry profitable? That depends on the extent to which the entry itself will depress price. Why does committed entry depress price? Because the entrant adds output to the market, which buyers will not absorb unless other sellers cut back or price falls. How much output does the entrant add to the market? At least the entrant’s “minimum viable scale,” i.e. the annual sales an entrant must make to guarantee that it can cover its fixed investment, given that prices will not exceed their premerger level if anticompetitive conduct is successfully deterred. “Minimum viable scale” is the entrant’s break-even annual sales at premerger prices, and typically expressed as a fraction of the industry’s annual sales.
The Guidelines point out that whether a given “minimum viable scale” is small enough to suggest that entry would likely deter or counteract the competitive problem depends upon the sales opportunities created by the feared anticompetitive effect and other factors such as the extent to which the entrant can divert sales from incumbents. The Guidelines even suggest a rule of thumb: when “minimum viable scale” substantially exceeds 5 percent of total market sales, we might question whether the entry plan will be profitable; but when “minimum viable scale” is markedly less than 5 percent of total market sales, profitability may seem more plausible.(36)
In demonstrating to a court that entry sufficient to cure the competitive problem would not be likely, the government will not necessarily try to prove “minimum viable scale” and “sales opportunities” quantitatively — nor should we. Yet even when we think judges and juries would not find mathematical calculations valuable, these concepts can structure the analysis and frame the testimony of those experienced in actual entry efforts in the industry. For example, industry witnesses who believe that an entrant would need a minimum market share to break even post-merger can be asked to explain why that share is so low or high. That will allow the court to see whether the entry plan the witnesses have in mind is plausible, by evaluating alternative views about the nature and the magnitude of the fixed expenditures required for entry. Such witnesses might also discuss how far prices are likely to fall following entry at a particular scale. The point is to focus the entry inquiry on the factors that determine whether committed entry in the post-merger market environment is likely to be profitable, and thus on whether entry is likely to deter or counteract the anticompetitive problem.
The 1992 Merger Guidelines highlight the way the analysis of committed entry differs from the analysis of unlimited uncommitted entry. The idea that ease of entry is a trump can apply to both, but it must be applied with particular care when entry is committed: committed entry is not a trump unless it would be timely, likely, and sufficient to deter or counteract the competitive harm from merger.(37)
Baker Hughes and Syufy
With this background, it is time to turn to the two problematic entry opinions written by the appellate courts in 1990, Baker Hughes and Syufy. I describe the opinions as problematic rather than the decisions to emphasize that the problem may be more with the language than the holdings. And let me acknowledge up front that it is not entirely fair for me to complain that the opinions are not clear about the distinction between committed and uncommitted entry when that distinction was not made in the prior case law and was not articulated in the Merger Guidelines until two years later. But it is mostly fair: after all, as I will describe later, the government worked hard in both cases to encourage the courts to recognize the importance of the distinction, though without benefit of the vocabulary and articulation in the 1992 Guidelines.
These two opinions encourage an analytical confusion in three related, perhaps equivalent, ways. First, they could be read to suggest that it is appropriate to apply the doctrine that ease of entry is a trump to evaluate committed entry cases without considering whether the committed entry would be timely, likely and sufficient. Doing so could lead courts to overestimate the ability of entrants to cure competitive problems from mergers. Second, they could be read to presume that any firm that could enter post-merger likely would enter, even when shown evidence that minimum viable scale is so large as to make it implausible that the entrant would receive a price high enough for profitability. Doing so could lead courts to overestimate the likelihood of entry. Third, they appear to analyze the height of entry barriers in the abstract, without connecting the entry analysis to the question of whether new competition solves a competitive problem — that is, without asking the sufficiency question. Doing so loses sight of the purpose of entry analysis: under the statute, entry is relevant only insofar as it bears on whether an acquisition’s effect “may be substantially to lessen competition, or to tend to create a monopoly.”
In both Baker Hughes and Syufy, the Justice Department accepted the legal proposition that entry could trump a prima facie case, but attempted to demonstrate that whatever entry there might be would not, or did not, cure the competitive problem. In both, the government had lost at trial and thus was required to accept the facts as found by the trial judge or else prove those findings “clearly erroneous.” The courts dismissed the government’s arguments out of hand, in both cases seeming to believe that Justice was ignoring the principle established in Waste Management and its own Guidelines, that entry could be a trump card. If this result represented simply a difference of opinion about the weight of the evidence, there would be little reason for concern. But the language the opinions use to explain the decisions could be read to stand for broad and troublesome propositions about the analysis of committed entry. And the Baker Hughes panel in particular is unusually influential: the D.C. Circuit’s decision was written by Clarence Thomas and joined by Ruth Ginsburg.(38)
Baker Hughes involved a challenge to a merger among firms selling a high-tech product, custom-built drilling machines for underground mining, with a very small U.S. market.(39) In describing the facts on entry, the appellate opinion emphasized that two firms had recently entered the U.S. market, Canadian firms were available to enter, and that volatility in market shares suggested that successful entry was possible.(40) Thus, the court arguably considered this an uncommitted entry case, with enough potential uncommitted entrants to make plausible the application of the trump. The court nevertheless did acknowledge “some facts suggesting difficulty of entry.(41) These were facts tending to suggest that entry would require significant sunk investments, and that such committed entry would not be likely to cure the competitive problem. In particular, the court recognized that the products at issue were custom-made, creating strong customer loyalty that an entrant would need to overcome. It also noted that buyers depended on assurances of product quality and reliable future service, assurances that an entrant could not easily provide. The Justice Department had noted these latter points in its briefs.(42) In short, the court arguably viewed the most plausible entry alternatives in Baker Hughes as uncommitted, while the government almost surely considered Baker Hughes to be a committed entry case.
In Baker Hughes, the government argued that, because it had shown high concentration and other elements supporting a presumption of anticompetitive effect, the defendants could avoid an injunction on the ground of easy entry only by a clear showing that entry would be “quick and effective” at undermining post-merger market power.(43) Perhaps because it saw the entry as uncommitted, the court seems to have missed what the government was arguing. Instead, the opinion claims that the government was trying to disavow its burden of proof of likely anticompetitive effects. It interpreted the government as claiming that the sole way to rebut an inference from high concentration is by showing ease of entry, when the government was merely setting forth an analysis that should be applied when entry is the subject of the rebuttal argument.(44) The court also rejected the government's insistence that entry must be “quick and effective” on the ground that this requirement overlooks the way potential competition, never resulting in actual entry, can nevertheless exert competitive pressure on a market.(45) Yet the court makes this point seemingly without noticing that potential competition cannot play this role unless entry, whether committed or uncommitted, would in fact be quick and effective.(46) In short, the court in effect viewed the government as having missed the point of Waste Management and its own Guidelines, that ease of entry is a trump. The Baker Hughes opinion is also noteworthy for its strong rhetoric. It reads like an exasperated effort to rein in a runaway agency thought to have willfully ignored the teaching of Waste Management.(47)
The court of appeals in Syufy also seemed to believe that the government could not read its own Guidelines(48) and would not recognize easy entry at high noon, let alone in a dark movie theater.(49) The case deals with a movie theater owner named Syufy who acquired all of his significant competitors in the Las Vegas movie theater market. The government charged that Syufy exercised monopsony power against distributors of first-run films in violation of both Sherman Act §2 and Clayton Act §7.(50) The alleged competitive problem was the reduction in prices paid to sellers of a key input, not higher ticket prices for moviegoers.(51) The trial court found for the defendant on the view that the case offered a “prime example” of the application of the legal proposition that ease of entry is a trump.(52)
One problem with the government’s case was that Syufy faced large-scale actual entry from the Roberts Company barely a year later. As a result, Syufy’s market share was declining substantially. It is worth pausing on this point. Although the appearence of actual entry, here as in Baker Hughes, may have encouraged the courts to think entry was easy, that inference does not necessarily follow.(53) In any event the circuit court dismissed, as inconsistent with the district court’s factual findings, government arguments that Roberts (and its successor, United Artists) was not an effective competitor,(54) and, as with the court in Baker Hughes, the Syufy panel arguably viewed the likely entry alternatives as uncommitted. It saw “a rough-and-tumble industry, marked by easy market access, fluid relationships with distributors, an ample and continuous supply of product, and a healthy and growing demand.(55) Accordingly, the Ninth Circuit upheld the trial court.
As with the D.C. Circuit panel that decided Baker Hughes, the Ninth Circuit panel that issued the Syufy opinion seems to have misunderstood what the government argued. The government presented an argument similar to the entry likelihood analysis for committed entry set forth in the 1992 Guidelines: it contended that entry at a scale large enough to achieve low costs would turn out to be unprofitable, because entry at that scale would depress market prices.(56) The Ninth Circuit’s opinion responds to this argument in two ways. The first response hides in a footnote: the court disputes the accuracy and meaning of evidence about market overcapacity and market growth for the assessment of the sales opportunities available to an entrant.(57) Had the opinion stopped here, dealing only with a difference of opinion about the weight of the evidence, the decision would not be remarkable.
But the opinion continues by arguing that the government rejects basic, long- established economic principles. The telling moment occurs when the opinion quotes government counsel at oral argument making the entry likelihood point about committed entry.(58) The court interpreted the government as endorsing what it calls “a shopworn argument we had thought long abandoned: that efficient, aggressive competition is itself a structural barrier to entry.(59) Perhaps because the court viewed entry as uncommitted, it did not recognize that the government was making a point about how to analyze committed entry. The government’s argument was about the scale necessary for an entrant to do business efficiently, and whether committed entry at that scale would be profitable; the court instead heard, and rejected, an argument about whether the incumbent was performing efficiently.
The Syufy opinion misses the committed entry point in part because it conceives of entry in terms of abstract, structural barriers, rather than connecting entry analysis to the question of whether new competition solves the competitive problem created by merger. Once the government “concedes that there are no structural barriers to entry into the market,(60) the court’s entry analysis is complete, as it is not appropriate to “speculate as to the details of a potential competitor’s performance.(61) And as with the Baker Hughes panel, the judges responsible for the Syufy decision appear to believe they needed to discipline an out-of-control government agency. The government’s complaint is, for the court, a counterproductive attack on an “efficient, vigorous, aggressive competitor,” not the protection of competition.(62)
The appellate decisions in Baker Hughes and Syufy may create a loophole in merger enforcement to the extent their language encourages courts to overestimate the ability of committed entrants to cure competitive problems from merger. The idea that ease of entry is a trump must be applied with care, particularly when committed entry is at stake. Unthinking application of the doctrine may encourage courts to analyze the height of entry barriers in the abstract, and not recognize that entry is relevant only to the extent that it cures the competitive problem at issue. As a result, it may lead courts to presume that a firm that could enter the market likely would find it profitable to do so. Yet when entry requires significant sunk investments, its profitability — entry likelihood — is a matter for analysis not presumption. A court that disregards the entry likelihood question — or one that presumes that examples of past entry are dispositive on the issue of the profitability of future entry that cures the competitive problem — may find itself wrongly allowing anticompetitive mergers to proceed.
Several merger cases in which entry played a role have been decided since the Baker Hughes and Syufy opinions were handed down, and the early returns suggest that some of the potential for analytical confusion is being realized. Three district courts considered whether a seller’s need to develop a reputation for quality makes committed entry unlikely, evaluating a view similar to the argument the government made in Baker Hughes. In two cases, Russell Stover(63) and Gillette,(64) the courts did what the Baker Hughes panel did — they noted instances of actual entry, deemed the barriers low, and allowed the merger. In the exception, the United Tote case,(65) the evidence that entry would not cure the competitive problem was overwhelming. The district court nevertheless felt obliged to acknowledge that Baker Hughes and Syufy seemed to point the other way before concluding that entry would not be sufficient.(66)
On the other hand, the FTC routinely asks the right questions when analyzing commmited entry — examining whether such entry would cure the competitive problem, and considering whether the scale needed for low cost entry would make it unprofitable.(67) But Commission opinions on this subject may be less influential than one would hope because they employ the entry barriers/entry impediments distinction developed in Echlin, which has not been adopted by the federal courts.(68) And last year a Ninth Circuit panel chose not to take its cue from the circuit court’s earlier Syufy decision. The Rebel Oil panel recognized that entry, though easy for some firms, may be insufficient to solve a competitive problem “if the market is unable to correct itself despite the entry of small rivals.(69) The court even cited the 1992 Merger Guidelines standard for testing committed entry — that it be “timely, likely, and sufficient” — as authority. But this decision does not control in merger reviews because the opinion was written in a monopolization case.(70)
When the next merger case turning on committed entry comes along, the government will face a difficult litigation problem. Two appellate panels have instructed the government to recognize that ease of entry is a trump, with language that appears to apply even when the entry involves significant sunk expenditures. Both panels wrote as if they needed to rein in an out-of-control Justice Department. The D.C. Circuit told the government directly that there is no basis in law for a requirement that entry be “quick and effective.” And the Baker Hughes and Syufy decisions remain the leading lower court precedents.
The response of the government enforcers may be misunderstood. Two years after Baker Hughes and Syufy, the Justice Department and Federal Trade Commission together issued revised Merger Guidelines that refrain from requiring that committed entry be “quick and effective” in favor of something sounding nearly identical: that it be “timely, likely, and sufficient.” A court already persuaded that government enforcers are out-of-control could read this language as a direct challenge to judicial authority by unrepentant agencies.
That reading would be a mistake. The government is not thumbing its nose at the courts. A principled basis for the government’s approach in the Merger Guidelines is that the courts that decided Baker Hughes and Syufy effectively concluded that uncommitted entry was easy on the facts of those cases. As understood by the courts, these were not committed entry decisions. Once a court concludes that entry was uncommitted and easy, it is appropriate to dismiss the case on the basis of the Waste Management doctrine that ease of entry is a trump. But when committed entry is at issue, the merger statute requires that the government and the courts evaluate whether entry would be profitable, in order to determine whether entry would likely cure the competitive problem.
Let me conclude with a prediction. After the Israelities were found worshipping a golden calf, Moses read them the Ten Commandments and returned them to the path toward the Promised Land. The 1992 Merger Guidelines propose only three commandments regarding committed entry: it must be timely, likely, and sufficient to cure the harm from merger. And the Supreme Court, in a decision that has not yet been written, will some day say so.
(*)The author is indebted to Stephen Calkins and Michael Wise.
(1)United States v. Baker Hughes, Inc., 908 F.2d 981 (D.C. Cir. 1990); United States v. Syufy Enterprises, 903 F.2d 659 (9th Cir. 1990).
(2)United States v. Columbia Steel Co., 334 U.S. 495, 510-11 (1948).
(3)Brown Shoe Co. v. United States, 370 U.S. 294, 325 n.42 (1962).
(4)United States v. Grinnell Corp., 384 U.S. 563, 571-73 (1966) (placing central station burglar alarm and fire alarm services in the same product market).
(5)United States v. E.I. duPont de Nemours Co., 351 U.S. 377, 404 (1966) (Cellophane).
(6)Compare United States v. Aluminum Co. of Am., 377 U.S. 271, 276-77 (1964) (Rome Cable) (insulated copper conductor and insulated aluminum conductor placed in separate markets because of insufficient demand substitutability) with Rome Cable at 285 (Stewart, J., dissenting) (district court’s broad market definition should be upheld based on both demand substitution and “complete manufacturing interchangeability”).
(7)L.G. Balfour v. FTC, 442 F.2d 1, 11 (7th Cir. 1971); Crown Zellerbach Corp. v. FTC, 296 F.2d 800, 812-13 (9th Cir. 1961), cert. denied, 370 U.S. 937 (1962).
(8)Telex Corp. v. International Business Mach. Corp., 510 F.2d 894, 914-19 (10th Cir.), cert. dismissed, 423 U.S. 802 (1975); Twin City Sportservice, Inc. v. Charles O. Finley & Co., 512 F.2d 1264, 1271-74 (9th Cir. 1975), aff’d after remand, 676 F.2d 1291 (9th Cir.) cert. denied, 459 U.S. 1009 (1982).
(9)Continental T.V., Inc. v. GTE Sylvania, Inc., 433 U.S. 36 (1977); Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc., 429 U.S. 477 (1977); Broadcast Music, Inc. v. Columbia Broadcasting System, Inc., 441 U.S. 1(1979) (BMI).
(10)510 F.2d at 916-17.
(11)Twin City Sportservice, 512 F.2d at 1272-73.
(12)Budd Co., 86 F.T.C. 518, 569-72 (1975); Yoder Bros., Inc. v. California-Florida Plant Corp., 537 F.2d 1347, 1368 (5th Cir. 1976), cert. denied, 429 U.S. 1094 (1977); R. Posner, Antitrust Law 127-28 (1976); Note, The Role of Supply Substitutability in Defining the Relevant Product Market, 65 Va. L. Rev. 129 (1979); Note, Potential Production: A Supply Side Approach for Relevant Product Market Definitions, 48 Fordham L. Rev. 1199 (1980); cf. Tenneco, Inc., 98 F.T.C. 464, 582 n.7 (1981) (recognizes principle), rev’d on other grounds, 689 F.2d 346 (2d Cir. 1982); Kaiser Aluminum & Chemical Corp., 93 F.T.C. 764 (1979) (considers production flexibility), aff’d, 652 F.2d 1324, 1330-32 (7th Cir. 1981) (although court refuses to consider production flexibility).
(13)United States v. General Dynamics Corp., 415 U.S. 486 (1974).
(14)Id. at 501-02.
(15)See, e.g., The Supreme Court, 1973 Term, 88 Harv. L. Rev. 41, 257 n.39 (1974) (observing simply that the Court’s 1974 merger decisions, which also included two bank merger cases, paid more attention than previous cases had to actual competitive effects).
(16)Telex justified its result by Grinnell. 510 F.2d at 919. Twin City relied on Columbia Steel and the Brown Shoe footnote. 512 F.2d. at 1271.
(17)See generally, Janusz A. Ordover & Jonathan B. Baker, Entry Analysis Under the 1992 Horizontal Merger Guidelines, 61 Antitrust L.J. 139 (1992).
(18)But some of the marketing expenditures used to extend the brand name to a new product category may have been sunk.
(19)U.S. Department of Justice and Federal Trade Commission, Horizontal Merger Guidelines, §1.41 (April 2, 1992) (1992 Merger Guidelines).
(20)United States v. Philadelphia Nat’l Bank, 374 U.S. 321 (1963).
(21)Philadelphia Nat’l Bank and General Dynamics remain the leading authorities structuring merger analysis under Clayton Act §7.
(22)The result was foreshadowed in United States v. M.P.M, Inc., 397 F.Supp 78, 92, 94 (D. Colo. 1975) (listing ease of entry as one of many reasons why an acquisition was not in violation of Section 7, despite creating concentration over the Philadelphia National Bank standard), and by dicta in United States v. Tracinda Investment Corp., 477 F.Supp. 1093, 1108 (C.D. Cal. 1979).
(23)United States v. Waste Management, Inc., 743 F.2d 976, 978, 983-84 (2d Cir. 1984).
(24)United States v. Waste Management, Inc., 588 F. Supp. 498, 513 (S.D.N.Y. 1983) (emphasis in original), rev’d, 743 F.2d 976 (2d Cir. 1984).
(25)United States v. Calmar Inc., 612 F.2d 1298, 1306-06 (D.N.J. 1985).
(26)Echlin Mfg. Co., 105 F.T.C. 479 (1985). A carburetor kit is a collection of parts used to “tune up” a defective carburetor. In dissent, Commissioner Bailey questioned the majority’s “single-minded focus” on the entry issue as “dispositive … where the prima facie case for antitrust concern … is as strong as it is here.” Id. at 493.
(27)Moreover, the 1982 Merger Guidelines had already stated the same proposition. U.S. Department of Justice Merger Guidelines, §III.B, 47 Fed. Reg. 28,493, 28,498 (1982).
(28)The Second Circuit may have been wrong to picture entry as uncommitted and unlimited in Waste Management. The district court had found that entry into trash collection was “relatively easy” for small scale startups, and that most such entrants had remained small or disappeared from the market. United States v. Waste Management, Inc., 588 F. Supp. 498, 513 (S.D.N.Y. 1983), rev’d, 743 F.2d 976 (2d Cir. 1984). Accordingly, it found “no showing of any circumstance, related to ease of entry or the trend of the business, which promises in and of itself to materially erode the competitive strength” of the merging firms, which together accounted for half of the business in the market. Id. at 513. In its brief seeking to uphold the district court on appeal, the Justice Department plausibly interpreted these statements, in light of the trial record and the lower court's conclusion that the merger would harm competition, as recognizing that neither new entrants nor small scale incumbents could successfully compete at the large scale necessary to undermine market power exercised by the merging market leaders. Brief of the United States at 49-53, United States v. Waste Management, Inc., 743 F.2d 976 (2d Cir., Nos. 83-6365, 84-6001) (April 23, 1984). But the Second Circuit read the lower court’s factual findings in an alternative way: as not distinguishing between large and small scale entry, and thus as concluding that entry would quickly eliminate “any anti-competitive impact of the merger.” 743 F.2d at 983. Thus, the Second Circuit understood the district court as finding that entry into trash collection was uncommitted and unlimited.
(29)Waste Management offered an opportunity, because the government saw entry as committed while the circuit court saw it as uncommitted. But the issue was not truly joined because of ambiguities in the district court opinion, as described in note 28, and the circuit court did not purport to deal with committed entry. The Echlin majority briefly addressed the significance of committed entry in dicta (having dismissed on the facts complaint counsel’s argument that entry would require significant sunk expenditures). But the discussion was no more than a place holder, offering little guidance. The opinion rejected the argument that sunk costs and scale economies “necessarily” create entry barriers without contending that they necessarily do not, and it did not consider whether sunk costs and scale economies could create an entry “impediment.” 105 F.T.C. at 486, 487-89.
(30)1992 Merger Guidelines, supra note 19, at §3.0.
(31)For a more extensive discussion, see generally, Ordover & Baker, supra note 17.
(32)1992 Merger Guidelines, supra note 19, at §3.0.
(33)Timeliness is related to sufficiency. “As a matter of theory, the timeliness requirement can be thought of as an intertemporal sufficiency requirement; entry will not be sufficient if it is delayed.” Ordover & Baker, supra note 17, at 145 n.23.
(34)It is appropriate to equate “likelihood” with “profitability” when considering the behavior of profit-maximizing firms. Here “profitability” is used in an economist’s sense, and thus takes into account the opportunity cost of capital to the firm.
(35)1992 Merger Guidelines, supra note 19, at §3.3.
(36)Id. A rule of thumb is just that. If other factors cited in the Guidelines are significant, these benchmarks may no longer be appropriate.
(37)Uncommitted entry is timely by definition, and it is likely because it does not require significant sunk expenditures. (It could be limited, though, in which case it may or may not be sufficient.)
(38)David Sentelle also served on the panel.
(39)Unit annual sales could be counted in three figures, while dollar sales totaled only a few millions. The trial judge reportedly took the government to task for having wasted resources on such a small target, admonishing the Antitrust Division to go after bigger fish, like the FTC was doing in challenging soft drink mergers. But there is no “small market” exemption from Section 7 that would permit a court to second guess the government’s exercise of prosecutorial discretion.
(40)United States v. Baker Hughes, Inc., 908 F.2d 981, 988-89 (D.C. Cir. 1990)
(41)” Id. at 989.
(42)Brief of the United States at 19-25, United States v. Baker Hughes, Inc., 908 F.2d 981 (D.C. Cir., No. 90-5060) (March 30, 1990). Justice also explained why it was not impressed by the examples of recent entry relied upon by the district court.
(43) Id. at 14; see id. at 16 (entry must be “sure, swift, and substantial” to be cognizable).
(44)Compare 908 F.2d at 983 (the government’s proposed standard is flawed because it “assumes that ease of entry by competitors is the only consideration relevant to a section 7 defendant’s rebuttal”) (emphasis in original), id. at 984-86 with Brief of the United States, supra note 42, at 13-14.
(45)908 F.2d at 988. The court also rejected the “quick and effective” test on the ground that it could not be applied in practice without effectively, and inappropriately, imposing on defendants the burden of proving that entry actually will occur. Id. at 987.
(46)But cf. 908 F.2d at 988 (“If the totality of a defendant’s evidence suggests that entry will be slow and ineffective, then the district court is unlikely to find the prima facie case rebutted. This is a far cry, however, from insisting that the defendant must invariably show that new competitors will enter quickly and effectively.”) (emphasis in original).
(47)The court “refers the government to its own merger guidelines,” 908 F.2d at 992 n.13 which the government thought it was explicating and implementing, not ignoring. The court highlights the alleged sarcasm of a rhetorical question, when the government asked whether “slow and ineffective entry” could rebut a prima facie case — even while conceding, in the next sentence, that the government’s prima facie case would stand unrebutted if defendant had shown only that entry would be slow and ineffective. Id. at 987-88. The panel twice describes itself as “at a loss to understand” the basis for a government argument, id. at 986, 987; points out what it sees as an inconsistency in the government’s brief, id. at 988; notes where the government “misses a crucial point,” id. at 988; and peppers the opinion with phrases like “no merit,” id. at 983, “devoid of support,” id. at 983, “flawed on the merits,” id. at 983, “misplaced” reliance, id. at 987, and “novel and unduly onerous,” id. at 987, to describe the government’s views, while endorsing the district court’s analysis as “fully consonant with precedent and logic.” Id. at 986.
(48) United States v. Syufy Enter., 903 F.2d 659, 666 n.11 (9th Cir. 1990).
(49)The opinion is of interest to puzzle aficionados, as Judge Kozinski reportedly buried over 200 movie titles in it.
(50)903 F.2d at 662 n.3.
(51) Id. at 663.
(52)United States v. Syufy Enter. 712 F.Supp. 1386, 1401 (N.D. Cal. 1989), aff’d, 903 F.2d 659 (9th Cir. 1990).
(53)Past committed entry was not necessarily undertaken with the expectation of receiving the premerger price in the long run, so does not prove that committed entry would be likely in response to the merger. Ordover & Baker, supra note 17, at 144 n.21.
(54)903 F.2d at 665 n.8, 669.
(55)” Id. at 667.
(56)Brief for the United States at 37-39, United States v. Syufy Enter., 903 F.2d 659 (9th Cir., No. 89-1575) (April 21, 1989).
(57)903 F.2d at 667 n13.
(58)Id. at 667-68. The opinion quotes the government counsel as arguing, “[E]ntry must hold some reasonable prospect of profitability for the entrant, or else the entrant will say … this is not an attractive market to enter … .[T]he reason is very clear. You have to compete effectively in this market. And witness after witness testified you would need to build anywhere from 12 to 24 theatres, which is a very expensive and time consuming proposition. And, you would then find yourself in a bidding war against Syufy.” (emphasis in original).
(59)” Id. at 667.
(60)” Id. at 666.
(61)” Id. at 667 n.13.
(62)Id. at 669. The government started at a disadvantage with the opinion’s author, Alex Kozinski. Judge Kozinski made his perspective about the role of the antitrust laws clear when he wrote: “In a free enterprise system decisions such as [merger agreements among competitors] … should be made by market actors responding to market forces, not by government bureaucrats pursuing their notions of how the market should operate.” Id. at 673. On the lookout for ways in which antitrust enforcement creates “a real danger of stifling competition and creativity in the marketplace,” id. Judge Kozinski thought he found one in the government’s analysis of the role of entry. Cf. Stephen Calkins & F. Warren-Boulton, The State of Antitrust in 1990 at 63 (unpublished paper precented at a Cato Institute conference, A Century of Antitrust: The Lessons, The Challenges, April 1990) (noting ways in which “the opinion exudes antipathy for merger enforcement”).
(63) Pennsylvania v. Russell Stover Candies, 1993-1 Trade Cas. (CCH) ¶70,224, at 70,093-94 (E.D. Pa. 1993).
(64) United States v. Gillette Co., 828 F.Supp. 78, 84-85 (D.D.C. 1993).
(65)United States v. United Tote, Inc., 768 F.Supp. 1064 (D. Del. 1991).
(66)Id. at 1075-76, 1079.
(67)Notably the 1994 Coke-Dr. Pepper opinion, Coca-Cola Co., Dkt. 9207, Trade Reg. Rep. (CCH) ¶ 23,625 (1994), but also Occidental Petroleum Corp., 115 F.T.C. 1010, 1241-43 (1992) and, in an opinion predating Baker Hughes and Syufy, B.F. Goodrich Co., 110 F.T.C. 207, 295-303 (1988).
(68)See, e.g., Coke-Dr. Pepper, at 23,342 (Commission combines the vocabulary of “barriers and impediments,” focusing on time and practicability, with Guidelines’ “timely, likely, and sufficient” test, focusing on constraints that would prevent “entry” from being competitively significant). The Commission defines an entry impediment as “any condition that necessarily delays entry ... and thus allows market power to be exercised in the interim,” and an entry barrier as “additional long-run costs that must be incurred by an entrant relative to the long-run costs faced by incumbent firms.” Id. (citing Echlin at 485-86).
(69)” Rebel Oil Co., Inc. v. Atlantic Richfield Co., 51 F. 3d 1421, 1440 (9th Cir. 1995).
(70)One circuit court that treated entry in a merger case in the wake of Baker Hughes and Syufy reached the unremarkable conclusion that legal barriers — hospital certificate of need laws — would preclude entry from solving the competitive problem. Federal Trade Commision v. University Health, Inc., 938 F.2d 1206 (11th Cir. 1991). On these facts, the court had no need to grapple with the details of entry analysis. Unfortunately, it analyzed the issue in terms of the height of barriers.