Restoring Balance in the Test for Exclusionary Conduct

• Publication year: 2015
• Author: By Thomas N. Dahdouh

RESTORING BALANCE IN THE TEST FOR EXCLUSIONARY CONDUCT

By Thomas N. Dahdouh¹

The Federal Trade Commission's recent monopolization cases against Intel² and Google,³ two strikingly different recent decisions by Courts of Appeals — Novell v. Microsoft⁴ and ZF Meritor v. Eaton Corp.,⁵ and a recent article by Professor Herbert Hovenkamp⁶ all demonstrate the ongoing tumult in Sherman Act Section Two monopolization theory. Courts have struggled for some time now with the vague and circular test for Sherman Act Section Two cases, first enunciated in United States v. Grinnell Corp.⁷ Unfortunately, the line between exclusionary conduct and legitimate conduct cannot be delineated easily. Faced with the difficulty in assessing conduct in a fact-intensive and nuanced manner, some courts have sought "one-size-fits-all" tests that would allow a court to dismiss summarily meritless cases without engaging in fact-intensive investigation. Indeed, the Supreme Court's decisions in Pacific Bell Telephone Co. v. linkLine Communications Inc.⁸ and Verizon Communications Inc. v. Law Offices of Curtis V. Trinko⁹ have certainly encouraged such efforts. The reality is that the near limitless variation of exclusionary conduct — and the serious potential that such conduct could seriously impair competition — makes any such effort to find a uniform test likely to lead to a seriously under-deterrent enforcement regime.

[Page 51]

Rather, the right enforcement approach must utilize a flexible balancing test. The District of Columbia Circuit Court, sitting en hanc in the Microsoft case, unanimously adopted the right overarching, rule-of-reason approach: a court must balance the anticompetitive effect of the exclusionary conduct against the procompetitive justification for the conduct.¹⁰ This article suggests a further addition to the overarching balancing test: namely, that the balancing test be adjusted depending on the type of exclusionary conduct at issue in the case. This article proposes that one would first (1) identify the type of conduct involved and slot it along a continuum from most suspect to least suspect; and then (2) adjust how demanding the "balancing test" is along the two key inquiries of the balancing test: (a) the level of evidence necessary to show that the asserted justification for the conduct is nonpretextual and cognizable (that is, procompetitive and efficiency-enhancing), and (b) the level of causal proof necessary to show that the conduct played a role in the creation or maintenance of the defendant's monopoly power.

This article will first explain this approach in more detail, before turning to discuss in more detail the allegations in Google, the Novell and ZF Meritor decisions, and Professor Hovenkamp's article on predatory pricing.

I. A CONDUCT-SPECIFIC BALANCING TEST

First, case-by-case adjudication by federal courts — and federal antitrust enforcement authorities — has developed implicit categories for exclusionary conduct from most to least suspect. The most suspect conduct includes the following:

• refusing to sell to a customer because the customer is also buying from a rival, as in Lorain Journal Co. v. United States;¹¹
• treating a customer differently because it is a rival or would-be rival, as in the FTC's first Intel case, FTC v. Intel Corp.;¹²
• "muscling" behavior — threats and tortious acts, including deceptive practices;¹³ and

[Page 52]

• entering into exclusives with customers that foreclose rivals from the most efficient distribution outlets, as in United States v. Dentsply International Inc.¹⁴

Nor far off would be conduct that falls in what has been termed "cheap exclusion." Cheap exclusion generally involves deception and tortious conduct.¹⁵ Exclusionary practices that are both inexpensive to undertake and incapable of yielding any cost-reducing efficiencies are "cheap" in both senses of that term and are most likely to appeal to a firm bent on maintaining or acquiring market power by anticompetitive means. Classic examples are instances of deception in the standard-setting process. For example, in Broadcom Corp. v. Qualcomm Inc.¹⁶ the court held that a standard-setting participant's false promise to license its patents that would read on the proposed standard could make out a monopolization claim. Another example is the Conwood case.¹⁷ In that case, the plaintiff had adduced evidence that the monopolist had engaged in a wide range of tortious conduct, including removing retail display racks with competitors' products without retailers' permission, providing misleading information to retailers about competing products and entering into exclusive agreements with retailers to exclude products.¹⁸

In the middle are near-exclusives, loyalty discounts or significant conditional discounts that lead to de facto exclusivity. These practices — commonly referred to as "minimum share discounts" — were part of the concerns that formed the basis of the second Intel case brought by the FTC.¹⁹ In Intel, the Commission found suspect the significance of the minimum share discounts involved — so-called "all unit" or first unit" discounts that were so large that relatively miniscule reductions in purchases could result in the loss of all the discounts.²⁰ Indeed, the OEM computer customers often

[Page 53]

needed these discounts simply to remain profitable in a particular market segment.²¹ The Commission also objected to the conditions placed on some of the discounts.²²

Also in the middle of the exclusionary continuum are claims that a manufacturer has created technological incompatibilities for interconnections with complementary products. In this second Intel case brought by the FTC,²³ the Commission indicated that Intel had crossed the line by generating incompatibilities in the "bus" — the connection that allowed the graphics chip to work together with the microprocessor — that would preclude competitors' previously-compatible graphics chips from working together with Intel's future microprocessors.²⁴ In that regard, the Commission alleged that Intel perceived the graphics chips manufacturers as a threat to Intel's microprocessor monopoly, as well as competition in the graphics chip market.²⁵

There is a fine line between issues relating to designing features and capabilities of a product that somehow adversely affect a rival or potential rival and incompatibilities in interconnections. The latter — suddenly ending a voluntary and presumably profitable course of enabling interconnections — may suggest that the change constitutes a deliberate effort to deter competition, particularly where the complementary product maker was poised to threaten the defendant's monopoly market. As described later in this article, the issues in its Google decision fell more in the first category, while the issues that the FTC grappled with in its 2009 Intel action sounded squarely in the second.

In the FTC's Google investigation,²⁶ certain companies alleged that Google's redesign of its search engine result page reduced the prominence in the display of their results. Design changes that have some adverse effect on other companies are, to my mind, much more difficult to challenge than technological incompatibilities. This is simply because valid business justifications are more likely for design

[Page 54]

changes that may make a product more desirable for consumers.²⁷ Consequently, design improvements are at the least-suspect end of the continuum. Courts have demonstrated some level of hostility to requiring a monopolist to design the features and capabilities of its own products to allow competitors to compete more effectively.²⁸

Following this continuum as outlined above can give a court a better sense of how to assess the particular conduct at issue. A court can fit the alleged conduct along this continuum and then assess it further. In the figure below, I have arrayed the different types of exclusionary conduct. A court could then employ a sliding scale for the balancing test focused on two key metrics: (1) assessing the asserted business justification for the conduct; and (2) assessing the causal link between the conduct and the creation or maintenance of the company's monopoly power.

Assessing the Business Justification for the Conduct: The D.C. Circuit in Microsoft described how to assess any procompetitive justification for the conduct. First, the justification must be "nonpretextual" — that is, there must be some basis in evidence to support the claimed justification — and, second, the justification must

[Page 55]

be cognizable — that is, it must provide a "great[] efficiency or enhanced consumer" benefit.²⁹ It is fairly easy to adjust these two key assessments and vary their strength, depending on how suspect the conduct is.

For conduct at the more suspect level, a greater quantum of evidence that the justification is non-pretextual may be necessary. For example, an adjudicator could require not just testimony supporting the claim of business justification, which may be self-serving, but also contemporaneous documents justifying the conduct at the time it is implemented. Similarly, to the extent that the conduct falls into the less-suspect end of the spectrum, a lower standard of proof may suffice.

On the issue of cognizability, one can similarly construct an array of different levels of proof. For example, at the more suspect level, one might require that the conduct be narrowly tailored to achieve the efficiency or enhanced consumer benefit. If less restrictive alternatives exist to achieve the same benefits, then the conduct can be condemned. As one moves toward the less suspect end of the spectrum, one could drop the requirement that the conduct be narrowly tailored. Even if there are other easily achievable methods to gain the efficiency goal, particular conduct might be acceptable, so long as it does in fact logically advance an efficiency-enhancing or beneficial-to-consumers goal.

Adjusting the Appropriate Level of Causal Proof: Causation is a...