Guideposts in the Analysis:

the Federal Trade Commission and
U.S. Department of Justice, Antitrust Division
Competitor Collaboration Guidelines

by

Susan S. DeSanti(1)

Director, Policy Planning
Federal Trade Commission

Houston Bar Association
Houston, Texas

December 7, 1999

Introduction

Thank you. I appreciate the invitation to talk about the recently released Antitrust Guidelines for Collaborations Among Competitors ("draft Guidelines" or "Guidelines"). As you know, the Guidelines were issued jointly by the Federal Trade Commission and the Antitrust Division of the Department of Justice ("the Agencies") in draft form. For those interested in submitting thoughts and suggestions, the time for submission of views has been extended and now runs through February 4, 2000.

I would like to begin today with a few general observations about the draft Guidelines, including the overall role of guidelines and how these draft Guidelines interact with other enforcement policy statements by the antitrust Agencies. I then will turn to the analytical framework in the draft Guidelines. I would be happy to answer questions at the end. As always, the views I express are my own and do not necessarily reflect those of the Commission or any Commissioner.

Observations about the Draft Guidelines

Let me start with a few observations. The fundamental goal of the antitrust Agencies' guidelines is to provide a framework that identifies the questions that the Agencies ask when analyzing particular transactions. In this sense, guidelines are an attempt by the Agencies to start everyone -- agency staff, the business community, antitrust counsel - on the same page. Guidelines, in short, are an effort to provide transparency. This is a particularly important effort in the context of agreements among competitors, where the case law has sometimes been ambiguous or even confusing, as we heard from participants in the Commission's 1995 hearings on global and innovation-based competition, who encouraged an effort to provide additional guidance on how the Agencies approach this area of the law.

Does that mean that agency guidelines identify every possible question? Of course not. Antitrust analysis, as we know, is highly fact intensive. Every case is different. This is particularly so when the subject is competitor collaborations, which may take forms as varied as incorporated joint venture entities, strategic alliances, or simple contractual arrangements, and which may involve diverse industries.

Does it mean that agency guidelines provide the answers to all of the questions that are identified? Again, the answer is of course not. Guidelines cannot and should not do so. Indeed, if they were to be too prescriptive and attempt to give all the answers, I would expect you to be appropriately alarmed. As the Supreme Court and the Agencies have approached competition issues in recent years, antitrust has become much too fact intensive an inquiry to allow any automatic answers to emerge from a certain set of questions.(2) Typically, the answer is "it depends," and what "it depends" on are the facts.

Where do the questions identified in the guidelines come from? They come from the Agencies' understanding and interpretation of the case law. The draft Guidelines reflect a mainstream understanding of the case law, except perhaps in a few instances -- which I will point out -- where, as a matter of agency discretion, the draft Guidelines arguably provide more leniency for or flexibility in the analysis of competitor collaborations than is found in the case law. The lesson from the now-rescinded Vertical Restraints Guidelines,(3) which attempted to push the law in a particular direction, is not to go beyond the boundaries of current case law, and these draft Guidelines remain safely within the lines.

One final, but very important, set of observations. The draft Guidelines are the first comprehensive set of guidelines issued jointly by both federal antitrust Agencies that address horizontal agreements among competitors. This is, in my view, one of the major contributions of the draft Guidelines. Unlike the Health Care Statements(4) and the Intellectual Property Guidelines,(5) the draft Guidelines apply broadly across many industries. Unlike the Merger Guidelines, the draft Guidelines apply not to a single type of transaction but rather to many types of agreements among competitors. They set out in one document the fundamental questions that are relevant across the board.

Two factors dictated the breadth and scope of these draft Guidelines. First, the Agencies recognize that in the information economy and other sectors of the economy as well, the types of "joint ventures" that occur among competitors no longer (if they ever did) stay within the contours of separate entities established by the venturers to carry out their joint activities, such as the paradigmatic joint venture to build and operate a factory in a "smokestack" industry. Rather, there are a multitude of possible arrangements, and the draft Guidelines reach agreements as diverse as R&D and marketing alliances, minority equity investments, and network arrangements. Guidelines that addressed only "entity" joint ventures would have quite limited applicability in today's markets. Second, the basic analytical framework that applies to all types of agreements among competitors is the same. Section 1 of the Sherman Act, Section 5 of the FTC Act, and the case law under those statutes provide the basic framework, which does not vary depending on whether an agreement results in the creation of a separate entity.

The breadth and scope of the draft Guidelines are also a source of limits on what they can accomplish, of course. Precisely because of their breadth and scope, the Guidelines do not specify exactly how much weight various factors will receive in the analysis of an agreement among competitors. The interaction among the relevant factors - such as the nature of the agreement, market share and market concentration, exclusivity or non-exclusivity, duration, etc. - varies too much from one transaction to the next to assign particular weights to certain factors. Instead, the draft Guidelines provide information on the factors that will be examined and the direction in which particular factors point - whether they likely increase or decrease competitive concerns. And, with the exception of the safety zones, the Guidelines do not set any market share or market concentration parameters that automatically either eliminate or presumptively increase competitive concern.(6) Again, this simply reflects the fact-driven nature of the antitrust analysis of competitor agreements; for example, a non-exclusive R&D collaboration among all firms in a relevant market might well be legal under the rule of reason.

So, a final question arises: Now that the Agencies have issued draft Guidelines that cover the broad area of competitor agreements, how do the new Guidelines interact with the Agencies' existing guidelines that treat, among other things, certain types of competitor agreements? The draft Guidelines are intended to work with, rather than supplant, the existing policy statements. Thus, the Health Care Statements should be consulted for antitrust questions involving health care, and the Intellectual Property Guidelines should be consulted for questions about licensing agreements for intellectual property protected by patent, copyright, and trade secret law. The Merger Guidelines should be consulted where competitor collaborations have effects identical to those that would arise if the participants merged in whole or in part; Section 1.3 of the draft Guidelines sets forth the criteria for when the Agencies analyze a competitor collaboration under the Merger Guidelines. In some cases, these more narrowly focused policy statements provide more specific information than would be possible in broad, generally applicable guidelines.(7) In general, though, all of the Agencies' policy statements are consistent and complementary.

The Draft Guidelines' Analysis

Now let me turn to the series of questions identified in the draft Guidelines. They begin with the focus of analysis: the relevant agreement.

The "Relevant Agreement"

The Guidelines define competitor collaborations as comprised of "a set of one or more agreements . . . ." Draft Guidelines, § 1.1. Section 2.3 explains:

In general, the Agencies assess the competitive effects of the overall collaboration and any individual agreement or set of agreements within the collaboration that may harm competition.

Whichever of these is under consideration -- the overall collaboration, an individual agreement, or a set of agreements -- is termed the "relevant agreement." Of course, where the competitive benefits or harms of two or more agreements are so intertwined that they cannot meaningfully be separated, the agreements must be analyzed together -- that is, as one set of agreements. See Section 2.3 and Appendix, Example 2.

Focus on the "relevant agreement" permits evaluation of individual agreements within a competitor collaboration.(8) Indeed, the Guidelines' central rule of reason inquiry asks "whether the relevant agreement likely harms competition by increasing the ability or incentive profitably to raise price above or reduce output, quality, service, or innovation below what likely would prevail in the absence of the relevant agreement." Id. §§ 1.2, 3.3 (emphasis added).

This agreement-by-agreement approach yields a more precise assessment of competitive effects than an analysis directed only at the collaboration as a whole. The draft Guidelines permit the Agencies to hone in on individual agreements to determine whether their anticompetitive harm is offset by their procompetitive benefits. In contrast, analysis based on the collaboration as a whole runs the risk of condemning an entire collaboration when only a single component causes anticompetitive harm. Alternatively, it might condone the anticompetitive harm from one agreement on the basis of procompetitive benefits deriving from another.

This ability to focus on individual agreements assumes particular significance in the context of network industries. For example, banks in credit card joint ventures traditionally have independently determined the interest rate that they charge consumers. If the banks in one such joint venture changed their rules and agreed to charge all users of the venture's card 23% annualized interest, the Guidelines could permit a challenge specifically focused on that agreement. The ability to address intra-network concerns of the type at issue in this example may be critically important, particularly where opportunities for inter-network competition are limited.

Time of Analysis

Competitive effects are assessed as of the time of possible harm to competition. The draft Guidelines, § 2.4, state the following general rule:

The Agencies assess the competitive effects of a relevant agreement as of the time of possible harm to competition, whether at formation of the collaboration or at a later date, as appropriate.

This recognizes that competitive effects may change as surrounding circumstances change. The Guidelines cite the example of an ATM network that bars its members from participating in other ATM networks. Appendix, Example 3. At the time the rule was adopted, the network was a small, fledgling effort, and the exclusivity requirement may have served a procompetitive purpose in ensuring sufficient business to get the new network off the ground. Years later, however, the network has grown greatly and now holds 60% of the ATM outlets in a relevant geographic market, but it continues to bar members from competing through membership in other networks. The analysis states that in assessing such an exclusivity rule, the Agencies would look to current circumstances, including the current market share and any other factors that suggest that the rule currently causes anticompetitive harm or currently provides procompetitive benefits. Id.

This reflects established agency practice in challenging agreements that become anticompetitive.(9) When competitors act in ways that restrict their rivalry, they run afoul of the antitrust laws from the time that their conduct impairs or becomes likely to impair competition. This neither changes antitrust rules in mid-stream nor punishes firms for their success. The same antitrust laws apply throughout, but when initially acceptable business conduct begins to impinge on those laws, its continuation becomes unlawful.

Indeed, in many situations the draft Guidelines' timing rule may be the only practical approach. Unlike mergers, competitor collaborations frequently go forward without Hart-Scott-Rodino filings, and the Agencies have no chance to conduct an investigation in advance. By the time an agreement comes to their attention, considerable time may have passed. Yet, reconstruction of initial market conditions may be wholly impractical. For example, consider the difficulty of retroactively delineating markets. If the Agencies were to start off by asking customers how they would have responded to a 5% price increase ten years ago, they would more likely get an incredulous laugh than any useful data.

Of course, rigidity in the face of changed circumstances can cause undesirable results, and the draft Guidelines temper their overall approach with two important caveats -- two caveats that go beyond anything compelled by the case law. First, the general rule is qualified as follows:

[A]n assessment after a collaboration has been formed is sensitive to the reasonable expectations of participants whose significant sunk cost investments in reliance on the relevant agreement were made before it became anticompetitive.

Draft Guidelines, § 2.4. This recognizes that a judicious exercise of prosecutorial discretion to account for reasonable expectations underlying sunk cost investment decisions based on initially lawful agreements may help to preserve a healthy investment environment.

Second, the general timing rule is superceded in the context of determining whether to challenge an agreement as per se unlawful. Under Section 3.2, the Agencies' per se inquiry considers whether practical, significantly less anticompetitive alternatives were available at the time the agreement was entered, rather than at the time of harm to competition. Except in unusual circumstances, imposing per se liability on the basis of alternatives not even existing when the relevant agreement was entered would seem too harsh.

The Per Se Rule

The draft Guidelines articulate a civil per se rule that involves two steps.(10) Step one identifies agreements that are potentially per se illegal, and step two identifies circumstances in which such agreements may escape per se condemnation and be reviewed under the rule of reason. Thus, for counseling purposes, absent hard-core cartel agreements, the first question is whether there are any agreements that are potentially per se illegal; the draft Guidelines note that "[t]ypically these are agreements not to compete on price or output." Section 3.2. The second question is whether any such agreements involve circumstances that justify review under the rule of reason. Id.

A two-step per se analysis in the civil context has long been recognized by the Supreme Court. At least since Broadcast Music, Inc. v. Columbia Broadcasting Systems,(11) the Supreme Court has considered justifications in determining whether an agreement is per se illegal. The Intellectual Property Guidelines and Health Care Statements also both apply a two-step formula.(12)

I would like to focus for a moment on the second step in the analysis. The draft Guidelines state that agreements that are "reasonably related to [an efficiency-enhancing integration of economic activity] and reasonably necessary to achieve its procompetitive benefits" will escape per se challenge. Draft Guidelines, § 3.2. Let me highlight a couple of the key aspects of step two.

First, step two requires an "efficiency-enhancing integration of economic activity." Although the concept of an efficiency-enhancing integration has long been applied in distinguishing per se illegal conduct from conduct subject to the rule of reason,(13) the challenge for the Guidelines is to construct a definition broad enough to capture today's legitimate contractual collaborations without including certain conduct that the antitrust laws historically have found to be anticompetitive. In short, the challenge is to articulate the line between efficiency-enhancing integration and cartel behavior, beyond simply stating the words themselves.

The draft Guidelines do so by describing an efficiency-enhancing integration both in terms of what it is and what it is not. "In an efficiency-enhancing integration," they explain, "participants collaborate to perform or cause to be performed . . . one or more business functions, such as production, distribution, or R&D, and thereby benefit or potentially benefit consumers by expanding output, reducing price, or enhancing quality, service, or innovation." Section 3.2. "Typically," participants in an efficiency-enhancing integration "combine, by contract or otherwise, significant capital, technology, or other complementary assets to achieve procompetitive benefits that the participants could not achieve separately." Id.

This "integration" concept encompasses almost any form of productive cooperation with potential for benefitting consumers. Thus, participants in an integration may jointly perform business functions, such as when two firms build and operate a factory together or conduct a joint advertising campaign. Alternatively, they may collaborate to enable or assist one of them to perform a business function, such as when one firm licenses R&D assets to another so that the latter can combine it with other assets and produce goods or services. Significantly, the draft Guidelines expressly recognize that participants may integrate by contract, combining assets in ways that enhance their individual business activities, such as through strategic alliances or network affiliations, without establishing a new legal entity. "Assets," according to its usual meaning, includes "any possession having value."(14)

The draft Guidelines then approach efficiency-enhancing integration from the opposite perspective, identifying collaborative activity that does not constitute efficiency-enhancing integration. First, the type of collaboration normally associated with naked cartel behavior is excluded: "The mere coordination of decisions on price, output, customers, territories, and the like is not integration . . . ."(15) Then, the draft Guidelines make plain that even some cost-reducing activity will not qualify: "cost savings without integration are not a basis for avoiding per se condemnation." Id. For example, a specialization agreement that merely allocates markets among participants but permits cost-savings by enabling each participant to concentrate on a single market is not integrative and will not survive per se scrutiny. Such agreements eliminate rather than enhance competition, and the draft Guidelines provide for challenging them under the per se rule. See Appendix, Example 5.

A second key aspect of step two in per se analysis is that the relevant agreement must be "reasonably necessary" to achieve procompetitive benefits that the integration promotes.(16) If equivalent or comparable procompetitive benefits could be achieved through practical, significantly less restrictive means, an agreement is not considered reasonably necessary. Section 3.2. The "reasonable necessity" requirement, though, is not intended to be overly stringent. As the draft Guidelines state, "an agreement may be reasonably necessary without being essential," and the Agencies will "not search for a theoretically less restrictive alternative that was not practical given the business realities." Id.

One final point on the per se rule. In assessing whether an agreement is reasonably necessary to achieve an efficiency-enhancing integration of economic activity, the Agencies perform only a limited factual inquiry. This is largely a plausibility standard. The draft Guideline state that the Agencies will look at whether "efficiencies from an agreement that are possible in theory are not plausible in the context of the particular collaboration" and otherwise screen out justifications that are plainly pretextual. Id.

The Rule of Reason

Section 3.3 contains the next series of questions, which apply to agreements that are not subject to per se challenge. This series of questions may be applied in a variety of ways that reflect the continuum of antitrust analysis that the rule of reason encompasses. The answers to certain questions may sometimes quickly identify agreements with no likely anticompetitive effects or those with likely anticompetitive effects. Depending on the factual circumstances, more or fewer questions may need answers.

There are two key concepts in this section. First, "[t]he Agencies focus on only those factors, and undertake only that factual inquiry, necessary to make a sound determination of the overall competitive effect of the relevant agreement. Ordinarily, no one factor is dispositive in the analysis." Section 3.3. Thus, the section provides for the possibility that a rule of reason analysis may be ended at any one of a number of points, once a sound determination of the overall competitive effect of the relevant agreement has been reached. As a practical matter, there are more points along the continuum where the analysis may be ended with a finding of no likely anticompetitive effects than the reverse. This is no surprise, since efficiencies must be taken into account before any conclusion of overall anticompetitive effect can be reached.

Second, the section is based on merger analysis as adjusted to account for the differences between mergers and competitor collaborations. The relevant factors, which I will discuss in more detail in the following sections, largely track the key elements of merger analysis as expressed in the Merger Guidelines: market share and concentration, theories of competitive harm, entry, and procompetitive justifications. Large differences appear when one considers the nature of the agreement, since competitor collaborations may involve diverse combinations of business functions, whereas mergers result in one entity performing all business functions in a particular relevant market. Thus, Section 3.31 of the draft Guidelines provides an illustrative list of types of competitor collaborations and competitive concerns and procompetitive benefits typically associated with each. This section assists in focusing the inquiry, since the nature of the agreement determines the types of anticompetitive harms that may be of concern. For other factors in the analysis, the degree of difference from merger analysis varies with the facts.

A Flexible Rule of Reason

The draft Guidelines present a flexible rule of reason analysis that "varies in focus and detail depending on the nature of the agreement and market circumstances." Section 3.3. As the Supreme Court stated in California Dental Ass'n,(17) "what is required . . . is an inquiry meet for the case, looking to the circumstances, details and logic of the restraint." Rule of reason analysis, thus, can be understood as a continuum.

At one end of the continuum, the analysis sometimes may be performed quickly either to exculpate or condemn an agreement.(18) The draft Guidelines provide that "if the nature of the agreement and the absence of market power together demonstrate the absence of anticompetitive harm," the Agencies do not challenge the agreement. Section 3.3. The draft Guidelines note that the absence of market power may be determined without defining a relevant market. "For example, if no market power is likely under any plausible market definition, it does not matter which one is correct." Id. at n.27.

The draft Guidelines alternatively provide: "[W]here the likelihood of anticompetitive harm is evident from the nature of the agreement, or anticompetitive harm has resulted from an agreement already in operation, then, absent overriding benefits that could offset the anticompetitive harm, the Agencies challenge such agreements without a detailed market analysis." Draft Guidelines, § 3.3. In some cases, an agreement already in operation may have caused anticompetitive harm. For example, in Indiana Federation of Dentists the Supreme Court found an actual restriction on output from the dentists' refusal to provide patients' x-rays.(19) In other cases, the nature of an agreement may evidence the likelihood of anticompetitive harm. As the Supreme Court stated in California Dental Association, sometimes an agreement "give[s] rise to an intuitively obvious inference of anticompetitive effect."(20)

What if a more detailed market analysis is necessary to assess accurately the competitive effect of an agreement? Under the Guidelines, the relevant factors include the nature of the agreement, market share and concentration, theories of competitive harm, entry, and procompetitive justifications.

Most of these factors are no doubt familiar to all of you from merger analysis. However, although the draft Guidelines adopt the structure of merger analysis, they modify it as necessary to account for the differences between mergers and collaborations. Moreover, although the Guidelines set out one possible order in which to conduct an analysis, a real-life inquiry is likely to be more flexible than any written structure might suggest. In practice, factual issues often are analyzed simultaneously rather than sequentially, so that a factor discussed relatively late in the Guidelines, such as ease of entry, may contribute early on to a conclusion that a given collaboration and the agreements of which it is comprised pose no likelihood of anticompetitive harm.

As noted previously, the first step is to identify the nature of the relevant agreement and competitive concerns and benefits that may be associated with it. Then, for more detailed analyses, the Agencies typically define relevant markets and calculate market shares and concentration. Sometimes it may be possible to assess competitive effects directly, without defining a particular relevant market, and sometimes, where the competitive concern is a reduction of R&D effort, it may be necessary to define an innovation market, but, by and large, the draft Guidelines look to Merger Guidelines principles for defining markets and identifying market participants.(21)

The draft Guidelines introduce an important adjustment when it comes time to assign the collaboration a market share. In the case of a merger, that process is relatively straightforward: the merged entity is treated as a combination of the merging parties, with one market share that results. However, competitor collaborations may differ in that sometimes, some elements of competition among the participants persist. In those circumstances, if the Guidelines attributed to the collaboration the same market share as if the participants had merged, that would overstate the potential competitive harm. Consequently, the draft Guidelines treat collaborations as having a range of possible market shares. At one extreme, where the participants are likely to remain independent factors in the relevant market, their shares should not be aggregated as subject to joint control; the Guidelines provide for a low end that is only the market share that the collaboration holds in isolation from its participants. At the other extreme, where the participants and the collaboration are likely to operate under joint control, the market shares would be aggregated much like a merger, i.e., the collaboration's market share would be the combined market share of the parents and the collaboration itself. Section 3.33.

The top end of the range provides the basis for the general safety zone found in Section 4 of the Guidelines. "The Agencies do not challenge a competitor collaboration when the market shares of the collaboration and its participants collectively account for no more than twenty percent of each relevant market in which competition would be affected." Section 4.2.(22) The Guidelines state that the safety zones "are not intended to discourage competitor collaborations that fall outside the safety zones," however. "The Agencies emphasize that competitor collaborations are not anticompetitive merely because they fall outside the safety zones." Section 4.1. For example, consistent with the shortened analysis outlined above, "if the nature of an agreement and the absence of market power together demonstrate the absence of anticompetitive harm," the Agencies do not challenge the agreement. Section 3.3.

Where that is not the case -- where, instead, "the nature of the agreement and market share and market concentration data reveal a likelihood of anticompetitive harm" -- Section 3.34 of the draft Guidelines identifies several factors that the Agencies find helpful in interpreting the extent to which the participants and the collaboration have the ability or incentive to continue to compete independent of each other in the relevant market, that is, to engage in what we have come to refer to as "insider competition." See Section 3.34.(23) Consideration of these factors may reduce or increase competitive concern.(24)

First, the Guidelines focus on "the extent to which the relevant agreement is non-exclusive in that participants are likely to continue to compete independently outside the collaboration in the market in which the collaboration operates." This factor should be familiar from the Health Care Statements, where the Agencies provide a broader safety zone for non-exclusive physician networks - whose members are available to compete independently or through competing networks - than for exclusive physician networks.(25) "In general," the Draft Guidelines explain, "competitive concern likely is reduced to the extent that participants actually have continued to compete, either through separate, independent business operations or through membership in other collaborations, or are permitted to do so." Section 3.34(a).

Second, the draft Guidelines take account of the nature and extent of each participant's financial interest in the collaboration. The thinking here is that a participant who holds a stake in the profits of the joint venture receives a lower net return from aggressive independent competition. "In general," the Guidelines tell us, "the greater the financial interest in the collaboration, the less likely is the participant to compete with the collaboration." Section 3.34(c). This has implications in a variety of contexts, ranging from situations where collaborations compete directly with their participants to settings where participants in strategic alliances solidify their relationship by taking minority equity stakes in one another.

A third factor identified in the Guidelines is the collaboration's organization and governance structure. Section 3.34(d). The collaboration may be set up as an independent decision maker in the relevant market, or it may be subject to participant control. The less the participants' control over the collaboration's price, output, and other competitively significant decisions, the more likely the collaboration will compete independently. Even assuming that participants control their collaboration, competitive concern is diminished to the extent that control is exercised independently, such as in a competitive rules joint venture, rather than jointly.

Duration may also affect incentives to compete. Collaborations with short time frames are likely to have less effect on the incentives of participants with long-run interests in strong, independent operations. Section 3.34(f). Other factors that may reduce competitive concerns include the participants' retention of control over assets needed to remain effective independent competitors and the adoption of appropriate safeguards to prevent anticompetitive information sharing. Sections 3.34(b), 3.34(e).  

In addition to asking about market shares and concentration and the ability and incentive of participants and the collaboration to continue to compete independent of each other in a relevant market, the Guidelines ask whether there is a likely anticompetitive story. They do this by cross-referencing the Merger Guidelines' discussions of adverse competitive effects. Section 3.3.

This helps to fill a void in the case law, which leaves important aspects of rule of reason analysis virtually uncharted territory. The Supreme Court's 1918 opinion in Chicago Board of Trade(26) listed numerous factors worthy of consideration. Eighty-one years later, the California Dental Association opinion told us that what is needed is "an enquiry meet for the case."(27) Between these cases the courts have provided strands of analysis, but have never really woven them together.

The draft Guidelines approach this task by looking to the same general theories of competitive harm articulated in the Merger Guidelines.(28) Harm may arise when a collaboration fosters express or tacit collusion among firms in the relevant market in a manner akin to the Merger Guidelines' coordinated interaction theories.(29) Alternatively, it may result from the combination of control and financial interests that fosters unified action by the collaboration, or by the collaboration and its participants, analogous to the Merger Guidelines' unilateral theories.(30) The draft Guidelines look to the nature of the relevant agreement and the market share and concentration data, interpreted in light of the potential for insider competition, and ask whether competitive harm under either of these general theories appears likely. Section 3.33.

If anticompetitive harm still appears likely -- that is, "[w]here the nature of the agreement and market share and market concentration data suggest a likelihood of anticompetitive harm that is not sufficiently mitigated by any continuing competition" - the draft Guidelines turn to an additional element familiar from merger analysis, but not so clearly ensconced in the rule of reason: they ask whether entry would be timely, likely and sufficient to deter or counteract the anticompetitive harm of concern. Section 3.35. Although here the draft Guidelines again draw upon principles of the Merger Guidelines, some qualifications are noted. The draft Guidelines explain that competitor collaborations may not provide the same inducements to enter or might not signal the presence of profit opportunities as clearly as do mergers. Id. For example, the likelihood of entry may be affected by what potential entrants believe about the relevant agreement's probable duration. If entry would take 18 months, potential entrants who think it likely that the collaboration will break down in roughly the same time frame lack the incentive to incur the sunk costs needed to enter. As a result, the anticompetitive harm would not be counteracted or deterred.

Finally, the draft Guidelines explain that "[i]f the Agencies conclude that the relevant agreement has caused, or is likely to cause, anticompetitive harm, they consider whether the agreement is reasonably necessary to achieve cognizable efficiencies." Section 3.36. Many of the concepts of the 1997 revisions to the Merger Guidelines are extended, almost verbatim, to the rule of reason. Cognizable efficiencies must be verifiable and potentially procompetitive, and the relevant agreement must be reasonably necessary for those efficiencies to be achieved. Sections 3.36(a), (b). The draft Guidelines make it clear that an agreement may be "reasonably necessary" without being essential, and that the Agencies will look only at "practical, significantly less restrictive" alternatives. Section 3.36(b). Finally, the draft Guidelines look to a determination of the relevant agreement's overall actual or likely effect on competition in the relevant market based on consideration of whether cognizable efficiencies likely would be sufficient to offset the potential of the agreement to harm consumers. Section 3.37.

In borrowing from merger analysis, however, the draft Guidelines once again tailor the analysis to reflect the fact that, although collaboration participants maintain their separate identities, joint conduct may be reasonably necessary to achieve the efficiencies of the collaboration. The Guidelines recognize that the continued independence of participants can pull a collaboration apart:

Collaborations sometimes include agreements to discourage any one participant from appropriating an undue share of the fruits of the collaboration or to align participants' incentives to encourage cooperation in achieving the efficiency goals of the collaboration.

Section 3.36(b). "The reasonable necessity of an agreement," the Guidelines explain, "may depend on whether it deters individual participants from undertaking free riding or other opportunistic conduct that could reduce significantly the ability of the collaboration to achieve cognizable efficiencies." Id. Indeed, the same agreement that heightens competitive concerns by limiting insider competition may be important for achieving procompetitive benefits. The draft Guidelines provide a framework that accounts for both consequences, each in the proper place, first in the discussion of anticompetitive harms and then in the discussion of procompetitive benefits.

Conclusion

I appreciate this opportunity to share with you some thoughts about the draft Guidelines. Now, let me remind you that it is your turn to share your thoughts and reactions with us. As I noted at the beginning, the time for submitting views has been extended to February 4, 2000, although we would be happy to get comments before then, if possible. I look forward to your comments, and I will be happy to take questions and comments now, as time permits.

Endnotes

1. I wish to thank Bill Cohen, Deputy Director, and Gary Zanfagna, Assistant Director, of Policy Planning for their outstanding and substantial assistance with this speech, as well as in the development of the draft Competitor Collaboration Guidelines.

2. Compare Broadcast Music, Inc. v. Columbia Broadcasting System, Inc., 441 U.S. 1 (1979) (further inquiry necessary to determine whether conduct that literally fixed prices was per se illegal) with United States v. Topco Associates, Inc., 405 U.S. 596 (1972) (agreement to divide market for Topco brand products held per se illegal without further inquiry). See also United States Dep't of Justice and Federal Trade Comm'n, Horizontal Merger Guidelines (1992, revised 1997), reprinted in 4 Trade Reg. Rep. (CCH) ¶ 13,104 ("Merger Guidelines") (Section 2.1, for example, calls for fact-intensive scrutiny in determining whether relevant market is conducive to collusion).

3. U.S. Dep't of Justice, Vertical Restraints Guidelines (1985), reprinted in 4 Trade Reg. Rep. (CCH) ¶ 13,105.

4. U.S. Dep't of Justice and Federal Trade Comm'n, Statements of Antitrust Enforcement Policy in Health Care (1996), reprinted in 4 Trade Reg. Rep. (CCH) ¶ 13,153 ("Health Care Statements").

5. U.S. Dep't of Justice and Federal Trade Comm'n, Antitrust Guidelines for the Licensing of Intellectual Property (1995), reprinted in 4 Trade Reg. Rep. (CCH) ¶ 13,132 ("Intellectual Property Guidelines").

6. Cf. Merger Guidelines, § 1.52 (stating presumptions of anticompetitive harm based on market concentration).

7. For example, Health Care Statement 9 on Multiprovider Networks provides that "[i]n some multiprovider networks, significant efficiencies may be achieved through agreement by the competing providers to share substantial financial risk for the services provided through the network [footnote omitted][,]" and states that "[i]n such cases, the setting of price would be integral to the network's use of such an arrangement and, therefore, would warrant evaluation under the rule of reason." Thus, for this type of arrangement, the agencies have provided a specific example of when a price agreement among health care providers would be "reasonably necessary to realize [significant] efficiencies [that benefit consumers]. [footnote omitted]"

8. The agreement-by-agreement approach has long been endorsed by case law. For example, in National Soc. of Prof'l. Eng'rs v. United States, 435 U.S. 679 (1978), the Supreme Court examined a specific ethical canon prohibiting competitive bidding, rather than the overall effects of an engineers' association. More recently, in California Dental Ass'n v. FTC, 119 S. Ct. 1604 (1999), the Court analyzed a professional association's advertising rules without mixing in consideration of procompetitive benefits that might have resulted from the "advantageous insurance and preferential financing arrangements" that the association facilitated or the lobbying, litigation, marketing and public relations activities that were conducted for its members. Id. at 1608, 1611.

9. The Intellectual Property Guidelines state an analysis focused on the time of harm to competition: "A determination by the Agencies that a restraint in a licensing arrangement qualifies for inclusion in the safety zone is based on the factual circumstances prevailing at the time of the conduct at issue." Id. § 4.3. The Health Care Statements have been interpreted similarly. See, e.g., DOJ Business Review Letter to Alan C. Nelsen (July 23, 1999) ("assuming that PPMG's membership continues to constitute 20 percent or fewer of the physicians in each physician specialty . . . it appears that PPMG would fall within the safety zone for exclusive physician network joint ventures described in Statement 8"). FTC advisory opinions routinely qualify their conclusions with the caveat that staff may reconsider the questions involved "if facts change significantly," and Antitrust Division business review letters state the absence of current intent to challenge transactions but reserve the right to take action if the transaction proves to be anticompetitive.

10. This two-step per se analysis applies only in the civil context. Conduct prosecuted criminally, which is solely within the jurisdiction of the Department of Justice, is not covered by the analysis. The Draft Guidelines state: "Because the courts conclusively presume . . . hard-core cartel agreements to be illegal, the Department of Justice treats them as such without inquiring into their claimed business purposes, anticompetitive harms, procompetitive benefits or overall competitive effects." Section 3.2.

11. 441 U.S. 1 (1979). In Broadcast Music, the Supreme Court explained that justifications are appropriately considered in order to determine whether an agreement is per se illegal. In holding that an agreement to jointly price a blanket license for music compositions, while literally price fixing, should nonetheless be assessed under the rule of reason, the Court explained:

"[P]rice fixing" is a short-hand way of describing certain categories of business behavior to which the per se rule has been held applicable . . . . Literalness is overly simplistic and often overbroad. When two partners set the price of their goods or services they are literally 'price fixing,' but they are not per se in violation of the Sherman Act. Thus, it is necessary to characterize the challenged conduct as falling within or without that category of behavior to which we apply the label "per se price fixing."

Id. at 9 (citation omitted). Since Broadcast Music, the Supreme Court repeatedly has recognized the role of justifications in determining whether an agreement is per se unlawful. See, e.g., Northwest Wholesale Stationers, Inc. v. Pacific Stationery & Printing Co., 472 U.S. 284, 294 (1985); NCAA, 468 U.S. at 100-01.

12. Intellectual Property Guidelines, § 3.4 ¶ 3; Health Care Statements 8.B.1 ¶ 1.

13. See, e.g., Maricopa, 457 U.S. at 339 n.7, 356-57 (finding no integration); Broadcast Music, 441 U.S. at 20 (finding integration); Health Care Statements 8.B.1, 9.A; Intellectual Property Guidelines, § 3.4.

14. The Oxford Dictionary and Thesaurus, Oxford University Press (American Edition 1996).

15. Draft Guidelines, § 3.2. At the same time, some integration is not "efficiency-enhancing." For example, a combination of cartel members' pricing departments in order to more accurately identify the monopoly price might be integrative, but it would not be "efficiency-enhancing" in the sense that it would promote procompetitive benefits.

16. Section 3.2. See Maricopa, 457 U.S. at 352-53 (not necessary for physicians to set maximum fee schedule for their own services when schedule set by insurers was a workable alternative); Broadcast Music, 441 U.S. at 20-21 (blanket license "necessary" for achieving integrative efficiencies and setting of price "necessary" for the blanket license).

17. 119 S. Ct. 1604, 1612-13, 1618 (1999).

18. NCAA, 468 U.S. 109 n.39 ("rule of reason can sometimes be applied in the twinkling of an eye") (quoting Phillip E. Areeda, The Rule of Reason" in Antitrust Analysis: General Issues 37-38 (Federal Judicial Center, June 1981)).

19. Indiana Fed'n of Dentists, 476 U.S. at 460-61.

20. California Dental Ass'n, 119 S. Ct. at 1617. See also Indiana Fed'n of Dentists, 476 U.S. at 459; NCAA, 468 U.S. at 106-10.

21. See Section 3.32(a). Because competitor collaborations are more likely than mergers to be evaluated after their implementation, the draft Guidelines give somewhat greater stress to the need to avoid the "Cellophane Trap," that is, to avoid defining markets too broadly when prevailing prices already reflect an exercise of market power. Id.

22. The safety zone does not apply, of course, to agreements that are per se illegal or that would be challenged without a detailed market analysis or to competitor collaborations to which a merger analysis is applied. Id.

23. See Michael S. McFalls, The Role and Assessment of Classical Market Power in Joint Venture Analysis, 66 Antitrust L.J. 651, 673-83 (1998); Steven C. Salop, When and How is it Proper for Competitors to Collude, in Osservatorio Giordano Ell'Amore Sui Rapporti Tra Diritto ed Economica del Centro Nazionale di Prevenzione e Difesa Sociale, The Value of Competition International Congress 218 (1992).

24. For example, closer examination of these factors may reveal that market shares should not be aggregated, or may reduce or increase concern about possible theories of competitive harm, such as collusion.

25. Health Care Statement 8.A.

26. Board of Trade of the City of Chicago v. United States, 246 U.S. 231, 238 (1918).

27. 119 Sup. Ct. at 1618.

28. Section 3.33 of the Guidelines points out that "[m]arket share and market concentration provide only a starting point for evaluating the competitive effect of the relevant agreement." In addition, the Agencies look to the factors outlined in Sections 2.1 and 2.2 of the Merger Guidelines, as well as those addressed in Section 1.52 of the Merger Guidelines, which addresses circumstances when market share or concentration data may overstate or understate likely competitive significance.

29. See Merger Guidelines, Section 2.1.

30. See Merger Guidelines, Section 2.2.