The Basic Structure of American Antitrust Law

When it comes to assuring low prices, high-quality goods and services, and product variety, there is no better regulator than market competition. Accordingly, the federal antitrust laws—chiefly, the Sherman and Clayton Acts—aim to promote vigorous competition among providers of goods and services. They do so by policing the two situations in which competition breaks down: monopoly (when there is a single dominant seller) and collusion (when nominal competitors agree not to compete).

The two main provisions of the Sherman Act correspond to these defects in competition. Section 1 forbids any “contract, combination … or conspiracy” in restraint of trade, which would cover collusion.[1] Section 2 addresses monopoly, making it illegal to “monopolize” or “attempt to monopolize” a market.[2] These provisions may be enforced by the government: the Federal Trade Commission (FTC)[3] and the Antitrust Division of the Department of Justice (DOJ). They are also enforceable by private plaintiffs who are “injured in [their] business or property” by antitrust violations.[4]

Because many antitrust violations, especially those involving collusion, occur in secret and cannot be successfully prosecuted, successful private plaintiffs are entitled to recover treble damages (i.e., three times the amount of loss occasioned by the antitrust violation).[5] Multiplying damages in this fashion aims to ensure “optimal deterrence” by creating appropriate incentives for potential violators, who are caught only a fraction of the time, to obey the law.

Implementing the antitrust laws has proven somewhat difficult. Much of the difficulty has stemmed from the breadth and generality of the statutory prohibitions.

• By its terms, Section 1 forbids all agreements that restrain trade, but virtually every commercial contract restrains trade: When I promise to sell something to you, I “restrain” myself from selling the thing to another. The U.S. Supreme Court has thus ruled that Section 1 forbids only agreements that “unreasonably” restrain trade.[6]
• Section 2’s prohibition of actual and attempted monopolization has been interpreted to require that the defendant (1) possess some degree of market power and (2) engage in “exclusionary” conduct,[7] but a great many pro-consumer business activities—e.g., price cuts and product enhancements—win business for the actor and thereby tend to “exclude” its rivals. Accordingly, courts now understand that a Section 2 violation requires “unreasonably” exclusionary conduct.

In most antitrust actions, courts must therefore assess the reasonableness of the conduct being challenged. They have generally done so by focusing on whether the challenged conduct enhances or reduces overall, quality-adjusted market output. If the practice at issue increases market output, benefiting consumers, the practice is reasonable; if it reduces market output, injuring consumers, the practice is unreasonable. Most of the time, reasonableness is assessed on a case-by-case basis, following a “rule of reason” that takes account of market features and the actual effect of the practice being challenged. For some practices, though, experience has consistently shown that the activity reduces market output, so an in-depth judicial evaluation seems unnecessary. Such practices are deemed to be “per se” unreasonable and thus illegal.

In the end, then, antitrust consists of a body of law that (1) attempts to ensure competition among market participants; (2) involves statutory prohibitions that are quite general and must be “fleshed out” by courts, which perform their task by assessing the likely effects of challenged conduct on overall market output; and (3) is highly attractive to private plaintiffs (and plaintiffs’ lawyers) seeking treble damages. Taken together, these features limit the benefits that antitrust confers on society.

The Limits of Antitrust and an Optimizing Approach

Antitrust is limited in its ability to enhance overall social welfare because its implementation is costly. Most obviously, there are significant costs in simply reaching a decision as to whether a particular business practice violates the antitrust laws. Because many business practices are “mixed bags” in that they may reduce competition on some fronts while enhancing it on others, the legality of a challenged practice generally turns on whether the practice, on the whole, is likely to enhance or reduce overall market output.

Answering that question, which ultimately determines whether the practice is “reasonable,” frequently requires a good deal of economic inquiry. Unless a practice has been classified as per se illegal, decision makers—both business planners assessing proposed conduct in advance and courts and enforcers evaluating it in retrospect—must often incur considerable costs in defining and evaluating the relevant market, studying the practice at issue, and assessing the conduct’s actual or likely effects on market output. Those are antitrust’s “decision costs.”

Antitrust adjudication will impose additional costs whenever the adjudicator reaches an incorrect liability decision. If the adjudicator wrongly fails to condemn an anticompetitive practice, the perpetrator(s) will gain “market power”—the ability to enhance profits by reducing output and hiking prices. The existence of such power tends to reduce social welfare as productive resources are allocated away from their highest and best ends.[8] On the other hand, if the adjudicator wrongly condemns a practice that enhances overall market output, consumers will be denied the benefit of enhanced production.

Such false condemnation of an efficient practice is likely to be even more costly than a false acquittal of an anticompetitive one. Whereas market power exists only in a particular market and tends to self-correct as new sellers enter that market in response to high prices, condemnation of an efficient business practice—the result of a false conviction—has economy-wide effects that cannot be corrected absent a court decision or statute overruling the decision that wrongly condemned the practice at issue. Taken together, the welfare losses from improper acquittals of anticompetitive practices and unwarranted convictions of output-enhancing practices comprise antitrust’s “error costs.”

Decision costs and the two types of error costs (from false acquittals and false convictions) are in inexorable tension. Simplifying a liability rule in an effort to reduce decision costs will tend to increase the incidence of wrong decisions and thereby enhance error costs. Making a rule more plaintiff-friendly in an effort to reduce false acquittals will increase false convictions. On the other hand, tweaking the rule in a defendant-friendly fashion may reduce false convictions but will increase false acquittals.

Making the liability rule more nuanced in an attempt to reduce both false acquittals and false condemnations will raise decision costs. As in a game of Whack-a-Mole, hammering down costs in one area causes them to spike elsewhere.

In light of this unhappy situation, then-Professor (now Judge) Frank H. Easterbrook, in an influential article entitled “The Limits of Antitrust,” recommended that antitrust tribunals give up trying to catch every anticompetitive act and endeavor instead to optimize antitrust. Specifically, Easterbrook argued, courts and enforcers should craft liability and procedural rules that minimize the sum of antitrust’s error and decision costs.[9] Taking that tack will ensure that the inherently limited antitrust laws generate as much social value as possible.

So how have the courts and enforcement agencies fared in harnessing Easterbrook’s insights to maximize antitrust’s social value? In recent years, the U.S. Supreme Court seems largely to have recognized antitrust’s limits and to have adopted rules consistent with Judge Easterbrook’s overarching policy prescription. The Court’s holdings are aimed at making it clearer to businesses what they can and cannot do, consistent with antitrust, and also providing substantial leeway for businesses to compete aggressively.

This approach helps consumers: When businesses have greater legal certainty, they have a stronger incentive to bring forth the new products and services that the public desires. Moreover, when companies spend less time treading cautiously to avoid antitrust investigations, their costs are lower, and this means that they can provide new offerings at lower prices than would otherwise be possible.

In recent years, however, the enforcement agencies have adopted various positions that are at odds with a “limits of antitrust” approach. This approach has spawned costly new uncertainty for American businesses and has prevented them (and thus consumers) from reaping the full benefits of the Supreme Court’s pragmatic and enlightened antitrust holdings.

The Roberts Court’s Respect for Antitrust’s Limits

Since John Roberts became its Chief Justice, the U.S. Supreme Court has taken quite an interest in antitrust.[10] Consideration of the Roberts Court’s seven decisions addressing vertical restraints of trade, exclusionary conduct, and antitrust enforcement reveals that each is consistent with a limits-of-antitrust approach, with several expressly seeking to minimize error and decision costs.[11]

Vertical Restraints. Vertical restraints of trade are trade-limiting agreements between economic actors at different stages in the distribution scheme (e.g., between a manufacturer and a retailer that carries its brand). One such restraint, “minimum resale price maintenance” or “RPM,” occurs when a manufacturer prohibits resellers of its brand, usually retailers, from selling below a certain price level.

The Supreme Court held in 1911 that RPM is always (“per se”) illegal as a form of “price fixing” that prevents price competition among retailers. Subsequent research, however, revealed that RPM often is used by manufacturers to incentivize distributors to provide point-of-sale services that consumers desire. Indeed, studies revealed that most instances of the practice increase rather than reduce market output. Thus, the per se rule generated substantial error costs by precluding many distribution contracts that could have raised consumer welfare.

Recognizing this, in 2007, the Supreme Court in Leegin overruled the old precedent and held that instances of RPM should be evaluated on a case-by-case basis under the antitrust “rule of reason.” By encouraging socially beneficial RPM contracts to proceed while maintaining courts’ ability to strike down truly harmful examples of RPM, the Leegin decision embodied a welfare-enhancing limits-of-antitrust approach to this important business practice.[12]

In its Independent Ink decision, the Court corrected another vertical restraints doctrine that threatened substantial error costs. Under long-standing precedent, tying—agreeing to sell some “tying” product (for example, a printer) on the condition that the buyer also purchase your “tied” product (for example, toner cartridges)—is per se illegal if (among other conditions) the defendant has market power (the ability to raise price and restrict output to less than competitive levels) over the tying product.[13]

The Court had suggested many years ago that the defendant’s mere possession of a patent on the tying product would be enough to establish market power over the tying product.[14] Such an understanding threatened to produce many false convictions because, in reality, most patents do not confer market power.[15] To prevent false convictions stemming from its improvident remark about patents and market power, the Court clarified in Independent Ink that mere possession of a patent on a tying product could not establish tying market power for purposes of a per se tying claim.

Both Leegin and Independent Ink, then, promise to reduce the error costs associated with antitrust regulation of vertical restraints.

Exclusionary Conduct. The Roberts Court has decided two cases involving conduct alleged to be unreasonably exclusionary and thus violative of Sherman Act Section 2. In Weyerhaeuser, the Court held that a plaintiff complaining of “predatory overbidding” by a rival buyer of inputs must demonstrate that (1) the defendant bid up input prices so high that the prices it charged for its output were below its own cost and (2) the defendant was likely to recoup its losses from below-cost pricing by underpaying for inputs once its predatory strategy eliminated rival input buyers.[16] The Court reasoned that its bright-line rule would capture most instances of anticompetitive bidding yet refrain from chilling legitimate bidding behavior. In other words, the two-part liability test would minimize social losses from false convictions and false acquittals while keeping administrative (decision) costs in check.

In LinkLine, the Court addressed “price squeezes” by “vertically integrated monopolists.”[17] In a price squeeze, a firm that has monopoly power over some input (e.g., aluminum ingots) but competes with others in selling a final product incorporating that input (e.g., rolled aluminum sheets) simultaneously raises its input price and lowers or holds constant its output price. Rival producers in the competitive output market find their profits “squeezed” because their costs rise, but they cannot increase their output prices without losing business to the vertically integrated input producer.

LinkLine held that a price squeeze does not constitute an independent violation of the Sherman Act.[18] To reach that conclusion, the Court relied on two precedents that had embraced a limits-of-antitrust approach. One was Trinko, which held that there is no general antitrust duty for a monopolist to deal with its rivals.[19] Any such duty, the Trinko Court reasoned, would (1) generate numerous and costly errors by encouraging collusion and reducing firms’ incentives to invest in economically beneficial facilities[20] and (2) entail high decision costs by “requir[ing] antitrust courts to act as central planners, identifying the proper price, quantity, and other terms of dealing—a role for which they are ill suited.”[21]

The other precedent upon which the LinkLine Court relied was Brooke Group, which held that a firm cannot be liable for predatory pricing unless it sets its price below its cost.[22] The Brooke Group Court emphasized that the liability rule it was imposing was not aimed at capturing every instance of anticompetitive low pricing, but rather was designed to condemn as many as possible without chilling procompetitive discounting practices and while keeping decision costs in check.[23]

Trinko and Brooke Group collectively foreclose liability based on a price squeeze. Trinko precludes liability based on the defendant’s hiking of input prices: If a firm has no general duty to deal with its rivals at all, then it certainly has no duty to sell to them at low prices. Brooke Group prevents liability for reducing output price to a non-predatory level (i.e., one that exceeds or equals the defendant’s cost). Given that Trinko and Brooke Group were themselves efforts to minimize the sum of error and decision costs, LinkLine’s holding—the logical result of combining the two precedents—similarly reflects such an effort.

Enforcement. The Roberts Court has decided three antitrust cases focusing not on substantive standards of liability for specific business practices but on antitrust enforcement generally. Again, each is consistent with a limits-of-antitrust approach.

In Twombly, the Court held that a plaintiff cannot plead the “agreement” (contract, combination, or conspiracy) element of a Section 1 claim merely by pointing to parallel behavior by the defendant and its alleged coconspirators and baldly asserting that the parties must have agreed to act in concert.[24] To survive a motion to dismiss for failure to plead an agreement, the Court held, a Section 1 plaintiff must make non-conclusory allegations “plausibly suggesting (not merely consistent with) agreement.”[25]

That holding recognized and attempted to address an inherent limitation in antitrust’s private enforcement scheme: The prospect of costly discovery and treble damages based on far-reaching economic harms often leads defendants to settle even meritless antitrust actions, and plaintiffs’ lawyers, well aware of defendants’ tendency to settle, have an incentive to conjure up bogus conspiracy claims any time they observe competitors acting in parallel fashion. By heightening the requirements for proceeding to discovery on an antitrust conspiracy claim, the Twombly Court reined in the costs associated with private enforcement of the antitrust laws.

The issue in the Court’s Credit Suisse decision was whether certain securities marketing practices that seemed to create an anticompetitive effect but were regulated by the Securities and Exchange Commission (SEC) could give rise to a private antitrust action.[26] In holding that they could not, the Supreme Court explicitly embraced a limits-of-antitrust approach, comparing the error costs of allowing versus not allowing an antitrust claim to proceed. Permitting the antitrust action, the Court reasoned, would entail a significant risk of false convictions, given that judges, unlike SEC officials, generally lack the ability to distinguish between desirable and undesirable securities marketing practices.[27] Moreover, the practices likely to be wrongly convicted or chilled tend to create significant social value.[28]

Taken together, these considerations suggest that error costs from allowing the action at issue could be quite large. On the other hand, any costs from not permitting the antitrust action to proceed would likely be small. Because the activity at issue was already regulated by the SEC, which has been directed “to take account of competitive considerations” in crafting its rules, there was little need to worry about social losses from falsely acquitting the behavior under the antitrust laws.[29] In the end, then, the Credit Suisse Court rested its holding on an explicit comparison of the error costs of false convictions versus false acquittals.

While the Roberts Court’s most recent enforcement decision expanded the scope of potential antitrust liability, it too is consistent with an effort to reduce the sum of antitrust’s error and decision costs. The issue in Phoebe Putney was whether a merger orchestrated by a local hospital authority was immune from federal antitrust liability.[30] Prior Supreme Court precedents had established that when a substate entity (e.g., some unit of local government) acts pursuant to authority granted by the state, its actions will be immune from federal antitrust law if the anticompetitive effect at issue was the “foreseeable result” of what the state authorized.[31]

The court of appeals had held that the loss of competition from the hospital merger being challenged was a foreseeable result of the state statute conferring general corporate powers, including the power to buy and sell hospitals, on the local hospital authority. In a unanimous reversal, the Supreme Court ruled that it is not enough for the state simply to grant general corporate powers, even acquisition authority. Instead, substate entities will be immune from antitrust liability only if either the authorizing legislature expressly contemplated the anticompetitive effect at issue or such effect was “the inherent, logical, or ordinary result of the exercise of the authority delegated by the state legislature.”[32]

This holding is consistent with an approach aimed at minimizing antitrust’s error and decision costs. Allowing a state’s granting of general corporate powers to immunize a substate entity from antitrust liability simply because the granted power could be exercised in an anticompetitive fashion creates a huge loophole that would acquit a significant number of truly anticompetitive acts. The Court’s holding avoids those errors without significantly increasing the likelihood of false convictions. After all, any state legislature that wishes to stay antitrust’s hand in order to pursue other laudable policies could express its intention to displace competition with the legislation delegating authority to the substate entity. Such intentional displacement of competition is likely to be uncommon, but it can be accomplished when desired.

In sum, the Roberts Court’s antitrust decisions addressing vertical restraints, exclusionary conduct, and antitrust enforcement have been uniformly consistent with an effort to minimize antitrust’s error and decision costs. We turn now to consider the degree to which the enforcement agencies have followed the Supreme Court’s lead.

The Enforcement Agencies’ Disregard for Antitrust’s Limits

During the time in which the Roberts Court has consistently crafted antitrust decisions that respect the law’s inherent limitations, the Federal Trade Commission and Department of Justice have moved in the opposite direction. In at least four areas, these agencies have taken actions that are likely to increase the sum of antitrust’s error and decision costs.

Exclusionary Conduct. The agencies’ insensitivity to the limits of antitrust is perhaps most evident in their rejection of enforcement guidelines on challenges to unilateral (single-firm) “exclusionary” (harmful) conduct. Recognizing the difficulty of distinguishing unreasonably exclusionary conduct from aggressive but legitimate competition, the FTC and DOJ set out in 2006 to provide some guidance on how the agencies would enforce the unilateral conduct provisions of Sherman Act Section 2. Over the course of a year, the agencies heard from 29 panels featuring 119 witnesses and considered substantial empirical evidence.[33]

The final Section 2 Report, released in September 2008 and comprising more than 200 pages, set forth principles to guide agency enforcement decisions in cases involving various categories of single-firm practices.[34] The report also addressed potentially exclusionary conduct not falling into one of the highlighted categories and thus not subject to a conduct-specific liability test. Such conduct, the report concluded, should be deemed unlawful under Section 2 only if its anticompetitive harms were shown to be substantially disproportionate to its procompetitive benefits.[35]

The Section 2 Report was closely attuned to the limits of antitrust. For each of the particular practices addressed, the report assessed why it is a competitive mixed bag—i.e., how it could occasion anticompetitive harm but might also create procompetitive benefits. The report then set forth liability rules designed to be both administrable and capable of condemning most anticompetitive instances of a practice while screening out those likely to create net benefits.

The report was particularly concerned about overdeterring output-enhancing behavior, reflecting a belief that market power is self-destructive (so that false conviction is of greater concern than false acquittal). In particular, the report sought to avoid overdeterrence by requiring that under Section 2 rule of reason analysis, a practice should be condemned only if it “disproportionately” impairs consumer welfare.

In the end, both enforcement agencies abandoned the Section 2 Report. The FTC never even signed on, with a majority of commissioners asserting that the report was deficient because it endorsed a limits-of-antitrust approach.[36] Specifically, the commission majority downplayed the risk of error, rejected the view that overdeterrence is of greater concern in antitrust than is underdeterrence, questioned the degree to which market power tends to be self-correcting, and discounted the value of administrable rules and screening devices.[37] Within a few months, the DOJ followed suit.[38]

The enforcement agencies’ abandonment of the Section 2 Report has liberated them to take aggressive enforcement actions against single-firm conduct that disadvantages a defendant’s rivals. It is doubtful, however, that such aggressive, competitor-focused enforcement will benefit consumers.

Consider, for example, the FTC’s late 2009 action against Intel Corp. The commission accused Intel of violating Section 2 by offering “loyalty rebates”[39] that would have passed muster under the guidelines in the Section 2 Report.[40] To settle the FTC action, Intel entered a consent decree in which it agreed not to give loyalty discounts in the future.[41] Surely, many of the forbidden discounts, which would have lowered consumer prices, would have been procompetitive.

Moreover, the FTC’s action reaches far beyond Intel. Because of the FTC action and subsequent settlement, similarly situated firms are likely to forgo their own loyalty discounts; they certainly have a fair response when purchasers of their products request such price cuts. To the extent that it increases the likelihood of enforcement actions like that in Intel, the agencies’ abandonment of the limits of antitrust–inspired Section 2 Report is likely to injure consumers.

Vertical Restraints. Despite empirical evidence demonstrating that most vertical trade restraints enhance rather than reduce market output,[42] the enforcement agencies remain hostile to such practices. With respect to minimum RPM, the FTC has indicated that it will continue to place a heavy burden on defendants. The commission has also taken a tough stance against exclusive dealing arrangements between manufacturers and retailers.

Minimum RPM. The commission set forth its position on minimum RPM in response to a request by women’s shoemaker Nine West to modify a consent decree the company had entered into when the practice was per se illegal. After the Leegin Court abrogated the per se rule, Nine West sought to void the parts of the consent order forbidding the company’s use of the practice.

The FTC ultimately agreed to loosen the constraints on Nine West, but in so doing, it set forth an evaluative approach that places a heavy burden on manufacturers considering—and thus discourages firms from implementing—RPM policies.[43] It therefore discourages the use of RPM arrangements, a perverse result in light of the evidence showing that such arrangements usually increase overall market output.

Exclusive Dealing. Recent FTC enforcement actions have displayed a similarly unwarranted hostility toward exclusive dealing arrangements. Exclusive dealing occurs when a seller (often a manufacturer) conditions the sale of its product on an agreement by the buyer (often a retailer) to purchase all of its requirements from that seller.[44]

Given that exclusive dealing frequently benefits consumers, the Supreme Court has long insisted that harm to competition—an actual or likely reduction in overall market output, not simply harm to an individual competitor—is necessary for antitrust liability based on the practice.[45] In recent months, however, the FTC has condemned exclusive dealing arrangements despite both a lack of evidence of anticompetitive harm and a procompetitive rationale for the arrangements under attack.

In its January 2014 McWane decision,[46] for example, the FTC condemned an exclusive dealing arrangement even though there was no direct evidence of consumer harm (i.e., higher prices or reduced output), but there was evidence that rivals of the manufacturer implementing the policy had entered the market and grown their market shares with the policy in place.[47] The record even demonstrated that the allegedly disadvantaged rival experienced the very same growth rate while the exclusive dealing policy was in place that it experienced in the periods before and after imposition of the policy.[48]

On top of all that, there was a procompetitive rationale for the policy: It protected McWane, which made a full line of domestic pipe fittings, from adverse cherry-picking by rivals that produced only the most popular, high-margin fittings. Such cherry-picking could have injured consumers by creating gaps in product availability.[49] The FTC’s condemnation of McWane’s exclusive dealing therefore thwarted a procompetitive benefit despite a lack of credible evidence of anticompetitive harm.

The commission’s recent Graco settlement similarly involved reflexive and unwarranted condemnation of exclusive dealing.[50] Graco, the leading manufacturer of “fast set equipment” (FSE) used to apply polyurethane coatings, acquired its two primary rivals in 2005 and 2008. The mergers extinguished competition in the North American FSE market, but a breakup of the companies was impracticable. The FTC therefore imposed restrictions on Graco’s conduct. The evidence showed that Graco had coerced FSE distributors into not carrying products of Graco’s rivals and had filed a questionable lawsuit against a rival supplier, causing FSE distributors to grow leery of that supplier and drop its products. Accordingly, the consent order prohibited Graco from engaging in distributor coercion and required dismissal of its questionable lawsuit.

But the order then went further. It prohibited Graco from entering into exclusive dealing contracts with distributors, and it placed limits on Graco’s freedom to grant loyalty discounts to distributors.[51] The problem with this is that there was no evidence that those last forbidden activities—exclusive dealing arrangements and loyalty discounts—contributed to the absence of competition in the FSE market; rather, they likely made the market more competitive.[52] By forbidding exclusive dealing and loyalty discounts, the commission’s consent order threatened to cause a consumer injury, and there was no reason to take such risk absent evidence that exclusive dealing had been used or was likely to be used in the future to create anticompetitive harm.

Intellectual Property Rights and Technology Standards. Back in the days when antitrust was less tightly moored to consumer welfare, the enforcement agencies were quick to find antitrust violations based on the exercise of intellectual property rights.[53] In recognition of the key role such rights play in furthering consumer welfare, however, the federal antitrust enforcement agencies eventually adopted a policy of treating intellectual property (IP) rights the same as conventional forms of property.[54]

The agencies proceeded from the premise that patent and other IP rights create incentives for innovation and its dissemination and commercialization by establishing enforceable property rights for the creators of new and useful products. Substandard protection for those rights (treating them as “second class property rights” that can arbitrarily be limited by government action) reduces investment in the creation of IP, which in turn slows the rate of innovation and retards the introduction of goods and services that consumers desire.

Unfortunately, however, federal antitrust enforcers now seem to be departing from the policy of supporting robust IP rights. In recent years, they have sought to use antitrust (including unfair competition law)[55] to constrain patent holders’ standard rights when their patents have been incorporated into technological standards.

The agencies’ purported goal in invoking antitrust in this context is to prevent anticompetitive “hold-up” by holders of “standard essential patents” (SEPs)—i.e., patents that a producer must license in order to utilize some technological standard (like the 4G standard for mobile telephones). “Commonly, businesses collaborate to establish [technology] standards by working through standard setting organizations (‘SSOs’) to develop a standard that all firms, regardless of whether they participate in the process, then can use in making products.”[56]

Once a technology standard has been widely adopted, SEP holders are in a good position to demand higher royalties from the producers implementing that standard.[57] To prevent such “hold-up” of vulnerable producers that have become locked in to a certain technology standard, SSOs frequently require that businesses active in an SSO must commit to licensing their patents that implicate features of a standard on “reasonable and non-discriminatory” (RAND) or “fair, reasonable, and non-discriminatory” (FRAND) terms.[58] The U.S. federal enforcement agencies have invoked antitrust to police two behaviors by SEP holders that have previously committed to grant licenses on F/RAND terms: suing for injunctive relief (rather than monetary damages)[59] and attempting to renegotiate license terms.[60]

While anticompetitive hold-up may be a legitimate concern when an SEP holder takes one of these actions, invoking antitrust in this context seems unwarranted. Patent and contract law already prevent anticompetitive hold-up here. Because the patent law standard for granting injunctions requires courts to consider the public interest when deciding whether to grant such relief,[61] it is wholly capable of precluding anticompetitive injunctions. Attempted license renegotiations, then, are subject to well-established contract doctrines that permit renegotiation when it is justified by legitimate commercial considerations, but not otherwise (especially if the effort to renegotiate is really just a hold-up attempt).[62] Given that patent and contract law can prevent anticompetitive hold-up from SEP holders’ injunction actions and attempted license renegotiations, invoking antitrust adds little, if any, social value.

On the other hand, using antitrust here does threaten significant error cost. There are legitimate reasons for an SEP holder that had previously made an F/RAND SSO commitment to engage in each of the “suspect” behaviors here. The SEP holder might appropriately seek injunctive relief rather than monetary damages against an infringer that was judgment-proof or had consistently and in bad faith rejected offered licenses in an attempt to gain bargaining leverage by forcing costly litigation. Or the SEP holder might legitimately try to renegotiate the case-specific meaning of F/RAND licensing terms in light of some market shift or other occurrence. Because SEP holders’ injunctive actions and attempts to renegotiate licenses are sometimes socially beneficial, the law should take care not to overdeter them.

For that reason, it makes little sense to invoke antitrust here. As explained above, successful antitrust actions result in treble damages to account for the fact that much antitrust misconduct (e.g., price-fixing) occurs in secret and is thus not successfully prosecuted; optimal deterrence requires a damages multiplier. For antitrust violations that do not occur in secret, however, the multiplier tends to create overdeterrence. That is the outcome here: Injunction actions and license renegotiations by SEP holders occur in the open, so applying a damages multiplier to account for the (nonexistent) difficulty of detection will overdeter.[63]

In sum, using antitrust to police hold-up stemming from injunctive actions or renegotiation efforts by SEP holders provides little marginal benefit (given that patent and contract law already police bad behavior here) while imposing significant marginal cost (given the likely overdeterrence resulting from potential antitrust liability). Respect for the limits of antitrust would call for the enforcement agencies to stay their hand in this context.

Merger Policy. The antitrust laws forbid business mergers that substantially lessen competition in a market.[64] To implement that prohibition, the FTC and DOJ review proposed mergers to assess their competitive effects. When an agency determines that a proposed merger is likely to lessen competition in a market, the agency may sue to enjoin the merger or may negotiate concessions (divestitures of certain business lines, etc.) designed to maintain competition. Two recent changes in the enforcement agencies’ merger review practices—reliance on the “gross” upward pricing pressure index (GUPPI) and embrace of conduct vs. structural remedies—suggest a lack of respect for the limits of antitrust.

Reliance on the GUPPI. The agencies have recently endorsed a difficult-to-administer, untested, and overly sensitive metric for identifying likely anticompetitive harm from “horizontal” mergers (i.e., mergers of competitors). Traditionally, the agencies assessed the likely competitive effects of planned mergers by first defining the market in which the merging parties participate and then evaluating a number of factors—the concentration of the market (i.e., the number of significant competitors), the ease with which new competitors could enter the market, the efficiencies the merger would create—to predict whether the merger would reduce overall market output.[65]

In revised Horizontal Merger Guidelines promulgated in 2010, the agencies explained that they may now determine that a merger is likely to be anticompetitive without first defining the market and considering how other firms would respond to the merger. Instead, a horizontal merger may be condemned solely on the basis that the merger is likely to create “upward pricing pressure” on some product sold by the merged firm. The notion is that mergers that unite closely competing brands of a product allow a firm that now controls the close substitutes to raise prices.[66]

Section 6 of the 2010 Horizontal Merger Guidelines calls for upward pricing pressure to be assessed without considering, at least initially, the degree to which the merger may create efficiencies that reduce the merged firm’s prices.[67] The guidelines thus contemplate the use of a “gross” upward pricing pressure index (GUPPI).[68] The GUPPI seeks to determine the likelihood, absent countervailing efficiencies, that the merged firm would seek to enhance its profits by raising the price of one of its competing products, knowing that some of the lost sales on that product will be diverted to the other.

Unfortunately, the GUPPI requires knowledge of complex economic data that are very hard to calculate in the real world.[69] Given the difficulty of figuring the GUPPI, mistakes are inevitable.

The agencies’ embrace of the GUPPI is also troubling because the metric has not been empirically verified as a means of identifying anticompetitive mergers. As economist Dennis Carlton has observed, “[U]se of the UPP as a merger screen is untested…. [T]here has been no empirical analysis that has been performed to validate its predictive value in assessing the competitive effects of mergers.”[70] This dearth of empirical evidence seems particularly troubling in that predicting the effects of mergers is no easy task.[71]

A third reason to be concerned about the agencies’ reliance on the GUPPI is that the index excludes a key consideration influencing the likelihood of adverse unilateral effects: cost efficiencies stemming from the merger.[72] Any upward pricing pressure resulting from the incentive of a merged firm to divert sales from one product to a higher-margin substitute should be balanced against merger-induced cost savings that would put downward pricing pressure on one or more of the merged company’s products. The 2010 guidelines, however, do not include an efficiencies credit in the initial calculation of upward pricing pressure. They instead relegate consideration of merger-induced efficiencies to a later analytical step under the standard efficiencies defense.[73]

In the end, the agencies’ reliance on the difficult-to-administer, empirically unverified, and inherently biased GUPPI is likely to generate many false condemnations of mergers that are, on the whole, beneficial.[74]

Conduct vs. Structural Remedies in Merger Review. Traditionally, when an enforcement agency concluded that a merger was likely to lessen competition, it imposed a “structural” remedy: either an order that the merger not proceed or a command that the parties divest some portion of the businesses to be merged. Enforcement of such a remedy was a simple matter; enforcers merely had to ensure that the parties did a single, discrete thing (i.e., cancel their merger plans altogether or first sell off some line of business). In recent years, though, the agencies have taken to approving mergers on the condition that the parties follow some set of detailed conduct rules.

The agencies’ embrace of a regulatory approach to merger remedies is most evident in DOJ’s 2011 Antitrust Division Policy Guide to Merger Remedies.[75] That document replaced DOJ’s 2004 Remedies Guide, which proclaimed that “[s]tructural remedies are preferred to conduct remedies in merger cases because they are relatively clean and certain, and generally avoid costly government entanglement in the market.”[76] The 2011 Remedies Guide eliminated (1) that statement, (2) a discussion of the limitations of conduct remedies, and (3) an assertion that behavioral remedies would be appropriate only in limited circumstances. It instead staked out a neutral position, stating that “[i]n certain factual circumstances, structural relief may be the best choice to preserve competition. In a different set of circumstances, behavioral relief may be the best choice.”[77]

Not surprisingly in light of the altered guidance, several of DOJ’s recent merger challenges have resulted in settlements involving detailed and significant regulation of the merged firm’s conduct. The settlements have featured mandatory licensing requirements, price regulation, compulsory arbitration of pricing disputes with recipients of mandated licenses, obligations to continue to develop and support certain products, establishment of informational firewalls between divisions of the merged companies, prohibitions on price and service discrimination among customers, and various reporting requirements.[78]

How this move toward regulatory merger remedies comports with a limits-of-antitrust perspective is somewhat unclear. On the one hand, if imposition of conduct remedies liberates procompetitive mergers that otherwise would have been barred outright, the trend toward greater use of such remedies may reduce overall error costs. By offering enforcers a less restrictive regulatory option—some middle ground between permitting the merger unconditionally and banning it or ordering divestment—conduct remedies could facilitate mergers that provide net benefits to consumers but raise concerns that cannot be addressed through divestiture. It appears, though, that conduct remedies are being used not to liberate otherwise banned mergers but to increase regulation of mergers that otherwise would have been approved unconditionally.

That is troubling, for conduct remedies present at least four difficulties from a limits-of-antitrust perspective.

• First and foremost, they are costly and hard to implement. They divert enforcers’ attention away from ferreting out anticompetitive conduct elsewhere in the economy and require managers of regulated firms to focus on appeasing regulators rather than on meeting their customers’ desires.
• Second, they may thwart procompetitive conduct by the regulated firm. When it comes to regulating how a firm interacts with its customers and rivals, it is extremely difficult to craft rules that will ban the bad without also precluding the good. For example, requiring a merged firm to charge all customers the same price (a commonly imposed conduct remedy) may make it hard for the firm to serve clients who impose higher costs.
• Third, conduct remedies tend to grow stale. Because competitive conditions are constantly changing, a conduct remedy that seems sensible when initially crafted may soon turn out to preclude beneficial business behavior.
• Finally, by transforming antitrust enforcers into regulatory agencies, conduct remedies invite wasteful lobbying and, ultimately, destructive agency capture.

In the end, these significant drawbacks likely outweigh any benefits from deterrence of anticompetitive conduct. As William Shughart and Diana Thomas have observed, “supervising compliance has been a backwater for the attorneys and economists employed by the federal antitrust agencies,” most of whom “do not want to be involved with ensuring compliance with court orders—job assignments that rarely make headlines.”[79] Accordingly, detailed conduct remedies often do not achieve their stated ends while still imposing significant costs. Their increased use seems inconsistent with a limits-of-antitrust approach.

Conclusion

The Roberts Court and the federal enforcement agencies have taken strikingly different stances on the limits of antitrust. The Roberts Court has generally respected them, crafting rules calculated to maximize antitrust’s social value in light of its inherent limitations. The agencies, by contrast, seem skeptical of the very existence of antitrust’s limits.

What is the reason for this divergence? An obvious explanation turns on the institutional features of federal courts versus agencies. Generalist judges, who regularly confront cases across the legal spectrum, know the limits of their expertise and, in light of their life tenure and limited opportunities for advancement, have no obvious need to expand their turf. Agency staff, by contrast, are constantly reminded of—and rewarded for—their specialized expertise, and they tend to gain both prestige and financial rewards as their authority expands. Their natural tendency is to expand the law’s reach.

Regardless of the cause of the diverging stances on the limits of antitrust, a couple of things are clear. First, recent enforcement agency policies are in severe tension with the philosophy that informs Supreme Court antitrust jurisprudence. Second, if the agencies do not reverse course, acknowledge antitrust’s limits, and seek to optimize the law in light of those limits, consumers and the competitive process will suffer. Let us hope that antitrust agency leaders heed this reality and adjust their enforcement policies accordingly.

—Thomas A. Lambert is Wall Chair of Corporate Law and Governance at the University of Missouri School of Law and co-author of Antitrust Law: Interpretation and Implementation (5th ed., Foundation Press, 2013).