Discussion to this point reveals the difficulty of articulating a standard for per se illegality. We want the standard to identify anticompetitive conduct with a high degree of accuracy. Although we do not demand absolute precision, it can be troubling if a standard would accord greatly different results to cases slightly to either side of a per se/rule-of-reason dividing line. We want the standard to be easily and quickly applied. We want the standard to give guidance to business, deterring unlawful conduct without interfering with beneficial activities. None of the standards thus far explored possesses all the ideal characteristics.
Further insights might be gained by stepping back from the specific articulations and assessing instead the general nature of the various methodologies. Approaches used in resolving per se/rule-of-reason issues have tended to fall into one of two categories. Some are self-standing in the sense that they provide a full and self-contained articulation of the per se/rule-of-reason dividing line. Most of the Supreme Court articulations have been self-standing. Other approaches are dependent. They take the established limits of per se conduct as their starting point and add to or subtract from those limits. The enforcement agencies' Health Care Statements 8.B.1 and 9.A, for example, are dependent; they analyze physician and multiprovider network joint ventures under the rule of reason if integration through the network is likely to produce significant efficiencies that benefit consumers and "any price agreements (or other agreements that would otherwise be per se illegal)" are reasonably necessary to realize those efficiencies. Each general category is discussed in turn.
Self-standing articulations have substantial benefits. By their very nature, they require that assessments reflect the full range of considerations underlying the per se rule. In keeping with antitrust's long tradition of evolving common law,(138) they permit both additions to and subtractions from conventional per se prohibitions. The former afford a natural outlet for the accumulation of experience with particular types of practices over time.(139) The latter implement BMI's rejection of literalness as a watchword for per se analysis. An obvious drawback to self-standing approaches has been the elusiveness of an ideal standard.
As a matter of day-to-day practice, self-standing per se rules often are framed in terms of functional categories of anticompetitive conduct. Thus, price fixing, market divisions, and certain types of group boycotts are termed per se illegal, subject, of course, to whatever role is assigned to procompetitive benefits. The problem is that the functional categories are not self-defining, as discussed supra in Section II.A.3. It then becomes necessary to refine the standard a level farther, such as by declaring per se unlawful all horizontal restraints that raise, lower, stabilize, or tamper with prices (again, subject to any reservations about procompetitive benefits).
This may work reasonably well over much of the spectrum of potential cases, e.g., those involving the raising, lowering, or stabilizing of prices. However, the categories tend to be quite unclear at their edges. The desire to proscribe as much plainly anticompetitive conduct as possible results in catch-all accretions (such as the prohibition against "tampering" with prices) that inherently lack the content needed for application. Moreover, focusing on categories of conduct raises the danger of the type of "formalistic line drawing" rejected by the Court, making attention to justifications and efficiencies critically important for ensuring that analysis is based on "demonstrable economic effect." Sylvania, 433 U.S. at 58-59. In addition, categorizations tend to generate a rigid per se/rule-of-reason dichotomy, yielding quite different results depending on which side of the blurry line a practice is deemed to fall.
The alternative has been to state standards that internalize the underlying economic considerations, looking to the economic significance or effect of the suspect conduct. This can be done very generally, as in the Supreme Court's summary proscription of conduct that is "plainly anticompetitive" or "manifestly anticompetitive," or that has a "pernicious effect on competition and lack of any redeeming virtue." If conduct came labeled as meeting these standards, they might yield desirable results, but for practical purposes, the standards lack specific content and can neither be easily applied nor reliably projected.
Other standards state their economic concerns with greater specificity. For example, we might ask whether a practice directly limits competition on price or output; whether it likely, or almost always, restricts competition and decreases output; or whether "on its face, it restricts output or restrains price competition,"(140) subject again to consideration of procompetitive benefits. This ensures a focus on economic consequences. It allows a continuum of analysis as the directness or seriousness of the restraint shades gradually from the traditional per se to the nearly per se portions of the spectrum. On the other hand, application of this approach requires economic judgment and sophistication. This raises a danger of muddying the waters at interior points well inside the traditional boundaries of the per se rule, with a concomitant loss of guidance, deterrence, and ease of administration.(141) Moreover, the sophisticated nature of the inquiry raises questions as to how deeply one need look in establishing the requisite limits on competition or likely price/output effects.(142)
Dependent tests permitting extension of per se treatment may be similarly divided between standards based on functional categories and those focused on economic content.
One approach relies upon analogies between the suspect conduct and established per se categories. For example, portions of the Commission's opinion in AIIC, slip op. at 37-40, grapple with whether rules establishing the length of interpreters' working days and the sizes of interpretation teams are more closely akin to per se unlawful price or output restraints or to standardization subject to the rule of reason. Reasoning by analogy permits some growth in per se coverage and ensures that results are grounded in tradition, but it may prove ambiguous in application. Absent mechanisms for measuring the closeness of a practice to the existing contours of per se categories, the process runs a danger of undue subjectivity in close cases.
In contrast, standards based on economic effect set more objective ground rules for extending per se coverage. If a new practice on its face restricts output and lacks offsetting procompetitive benefits, or if it is now understood that a previously considered practice almost always will reduce competition, the conduct can be condemned under the core rationale of the per se rule. Using a similar approach, but taking a closer look at asserted justifications, the Stepwise analysis permits extension of per se principles to new ground.(143) The focus on restraints that directly limit competition on price or output seeks to identify conduct that, absent offsetting benefits, warrants summary condemnation. If efficiencies are absent, the practice is condemned without demonstration of market power or specific anticompetitive effects.
Standards based on functional categories do not lend themselves to ready contraction of per se coverage. BMI highlighted their limitations, stating that practices that literally fix prices are not per se unlawful unless they fall within "that category of behavior to which we apply the label 'per se price fixing.'" 441 U.S. at 9. This provided little guidance by itself, and a shift in focus to the economic role of blanket licenses became necessary in order to advance the inquiry.(144)
Formulations with greater economic content typically have looked directly for the presence of procompetitive justifications or have employed ancillary restraints analysis. By their nature these approaches cut back on otherwise per se territory. The first derives from the Supreme Court's "effects-based" standards for per se illegality; the second is the progeny of Addyston Pipe, Rothery, and numerous other appellate cases and commentaries. The approaches may coincide or diverge depending on (i) whether the analysis is framed in terms of "relationship" or "necessity" and (ii) whether restraints are expected to be ancillary to an integration or to an efficiency.
The choice between "necessity" and "relationship" standards in many ways is a choice between accuracy and ease of application/predictability. In theory, if conduct is known to be unnecessary for achieving a procompetitive benefit, there is no need to revisit a presumption that it causes competitive harm. In practice, however, it often may be difficult to determine whether suspect conduct is necessary without a potentially complex inquiry into less anticompetitive alternatives. This would run counter to the precept that per se rules should be clear and readily applied. Some commentators have suggested that the per se rule's balance of accuracy and predictability might best served by an approach that (i) retains per se treatment when otherwise-per-se-unlawful conduct is clearly unnecessary to achieving the alleged benefits, such as when less anticompetitive alternatives are "obvious" or "clear," and (ii) invokes the rule of reason when the conduct is either (a) necessary for achieving the benefits or (b) related to those benefits and determining necessity would require complex inquiry.(145)
Many of the factors pertinent to choosing between an integration-based and an efficiency standard have already been identified supra in Section II.B.3. That discussion focused on integration's role as a proxy for efficiencies and tentatively concluded that (i) the utility of integration-based standards is significantly affected by the breadth of their definition; (ii) the narrower, asset-based definitions of integration may be the most easily applied; (iii) a broader definition of integration, requiring only coordination of functions and operations, may sometimes be needed to reach the full range of procompetitive efficiencies; and (iv) reliance on most concepts of integration tends to exclude the types of cost savings that antitrust law traditionally has not recognized as procompetitive. Section II.B.5 supra discussed the further consideration that some restraints may be related to an integration without being related to its efficiencies.(146)
Apart from any role as a proxy for efficiencies, integration may have separate significance by contributing to a practice's potential anticompetitive effects. Thus, some have argued that the greater the integration among a joint venture's parents, the less their ability or incentive to compete with each other or with their joint venture:Integration reduces competition among joint venture partners by combining into one entity the types of resources (e.g., intellectual property, production facilities, and capital resources) which the partners could have used to compete against each other. Furthermore, to the extent they have committed substantial resources, the partners will be less inclined to compete against a joint venture.(147)
In addition, it has been argued, greater integration among parents makes it more difficult to terminate their collaboration, so that any anticompetitive effects are more likely to be long-lasting.(148) Nonetheless, integration may retain viability for the purpose of drawing per se/rule-of-reason distinctions. The objective in that context is not to draw conclusions as to the net competitive effect of suspect conduct, but rather to ascertain whether there is sufficient likelihood of potentially offsetting efficiencies to warrant a closer look. To the extent that integration provides an accurate, readily applied proxy for likely procompetitive efficiencies, it serves that limited objective irrespective of whether it also suggests more acute competitive concerns.
Both the effects-based and ancillary restraints analyses set ground rules for considering procompetitive justifications, in keeping with modern concepts of per se analysis. Depending on how they are phrased, they yield varying points of equipoise in balancing accuracy against ease and clarity of implementation.
Since the Supreme Court's 1979 decision in BMI, it has been clear that per se analysis of competitor collaborations can no longer rely upon rigid categories and literal labels, but must instead take appropriate cognizance of conduct's likely economic effects. That does not mean that per se analysis has been replaced by the rule of reason or even that in every case it has been broadened. BMI itself recognized that per se analysis should continue to be conducted quickly and that benefits of clarity and deterrence may only be achieved by allowing a certain margin for error. Nonetheless, in determining whether it facially appears that conduct will always or almost always prove anticompetitive, the courts have exhibited an increasing willingness to consider procompetitive justifications.
This paper has considered various ground rules for opening per se analysis to justifications. Where a price or output restraint or a market allocation lacks any arguable competitive benefits, it can still be condemned out of hand, and "quick look per se" techniques have been suggested for similarly condemning arrangements that are effectively shams. When justifications appear more likely to have substance, the courts have required more detailed assessments of the conduct under the rule of reason. The degree of detail, however, may vary, depending on the level of scrutiny necessary for understanding a restraint's competitive effect. When anticompetitive consequences may be presumed from the very nature of the conduct, truncation techniques subject proffered justifications to review on the evidence, but permit quick condemnation if those justifications fail.
Precisely how to frame the governing standards remains in question. This paper has considered the pros and cons of several variants, including effects-based and ancillary restraints analyses, necessity and relationship formulations, and integration and efficiency tests. It has also examined possibilities for achieving a greater synthesis of per se and rule-of-reason thinking. No one standard appears perfect, and the debate may ultimately reduce to a policy choice seeking the best blend of accuracy, clarity, and litigation efficiency, suitable for quickly condemning practices that almost always prove anticompetitive and for deterring harmful conduct without inhibiting beneficial collaborative efforts.
138. See Khan, 1997 U.S. LEXIS 6705, at *29-30 (emphasizing that in drawing lines of per se illegality, as in other aspects of antitrust law, there is an interest in "recognizing and adapting to changed circumstances and the lessons of accumulated experience"); Business Electronics, 485 U.S. at 732.
139. As discussed supra in Section II.A.2, the courts have hesitated to apply per se rules when experience was lacking. The subsequent accumulation of experience would seem to remove that bar: "A per se category of violation may emerge as courts gain familiarity with the almost invariably untoward effects of a particular practice across economic actors and circumstances." CDA I, slip op. at 15; see also Klein, supra note 128, at 7. However, despite acknowledging the Commission's enhanced experience with advertising restrictions, the Ninth Circuit rejected application of the per se rule to the specific price advertising restraints at hand. CDA II, 1997 U.S. App. LEXIS 28882, at *17-18.
140. The first of these formulations is an element of the Stepwise Approach. The second derives from BMI's per se articulation, 441 U.S. at 20, and from Mass. Board's definition of "inherently suspect," 110 F.T.C. at 604. The third formulation was offered in the Brief for the United States as Amicus Curiae in NCAA at 12-13 (Jan. 1984) ("U.S. NCAA Brief").
141. For example, commentators have noted that a test focused on whether restraints "directly" affect competition and commerce was utilized in the early years of antitrust analysis with unsatisfactory results. See Ross, supra note 30, at 119-21, 127 (describing the Supreme Court's "directness" standard for defining restraints of trade subject to Section 1 condemnation, as applied in Joint Traffic Association and other early cases).
142. Typically, a quick look is thought sufficient. Thus, both BMI and the U.S. NCAA Brief at 12-13 speak in terms of a "facial" review. See Areeda, supra note 6, at ¶ 1511 (plaintiff's prima facie showing made "by argument"). Yet others might require a more searching inquiry. See James L. Langenfeld and Louis Silvia, Federal Trade Commission Horizontal Restraint Cases: An Economic Perspective, 61 Antitrust L.J. 653, 680 (1993) (requiring both a "persuasive theory" of how a restraint could cause competitive harm and "some credible evidence that is consistent with the anticompetitive theory" before applying Mass. Board's reformulation of BMI's "restrict competition and decrease output" standard).
143. The Stepwise Approach explicitly leaves established per se proscriptions in place and builds outward from there. See Klein, supra note 128, at 6-7.
144. Similarly, in the group boycott context, the Court found that "[e]xactly what types of activity fall within the forbidden category is . . . far from certain" and turned to economic requirements based on the presence of "market power or exclusive access to an element essential to effective competition" for ensuring a sufficient likelihood of anticompetitive effect. Northwest Wholesale Stationers, 472 U.S. at 294, 296.
145. See, e.g., Ross, supra note 30, at 142-43.
146. For this reason, some commentators have argued that an integration-based standard may detract from the per se rule's effectiveness by invoking the rule of reason too readily. See Section II.B.5 supra.
147. Thomas A. Piraino, Jr., Comment on Issues Relating to Joint Venture Project 5-6 (June 12, 1997). See Statement of Ernest Gellhorn 12 (June 30, 1997).
148.See Testimony of Ernest Gellhorn, Tr. 63-64 (June 30, 1997).