polygramopinion

1 OPINION OF THE COMMISSION BY MURIS, Chairman: INTRODUCTION Nessun Dorma! – None must sleep! This Puccini aria, sung by tenor Luciano Pavarotti in the recording at the heart of our case, announces the edict of the Chinese princess Turandot that no one in Peking may sleep until she solves her problem. The princess has made a bad judgment – agreeing to marry the first suitor who, at peril of death, can answer three riddles. Although this plan once had served her purposes, someone has now answered the riddles, and Turandot is encumbered with a product she neither wants nor can market. She grasps at one last chance to stop the wedding, by guessing the name of the suitor, and will stop at nothing to obtain the information. Our story takes place not on the opera stage, but in the business world of operatic recordings. The drama is not so stirring, and no one loses his head, at least not literally. The story is troubling, nonetheless. Two recording companies agree to form a joint venture to market a new recording, by three of the world’s foremost singers, and to split the costs and profits. By itself, such an agreement, even by competitors, is often beneficial, because it helps bring a new product to market. Here, however, the story turns dark when it becomes apparent that the new recording will repeat much of the repertoire of existing recordings, dimin- ishing its marketing potential and worrying the recording companies. While other businesses might have worked harder to develop an improved or more distinctive product to attract greater consumer interest, our protagonists chose another route. They agreed to restrict their marketing of competing products that they respective- ly controlled – products that were clearly outside the joint venture they had formed. They imposed a moratorium on discounting and promotion of those recordings that might otherwise siphon off sales of the new product. We now consider whether such an agreement unreasonably restrains trade in violation of the antitrust laws. We conclude that it does. 1 Comprehensive recent treatments of the relevant case law and commentary appear in ABA Antitrust Section, Monograph No. 23, The Rule of Reason (1999); VII Phillip E. Areeda & Herbert Hovenkamp, Antitrust Law ¶¶ 1500-12 (2d ed. 2003); Symposium: The Future Course of the Rule of Reason, 68 Antitrust L.J. 331 (2000). 2 Major FTC contributions to horizontal restraints jurisprudence include Pacific States Paper Trade Ass’n, 7 F.T.C. 155 (1923) (condemning agreement by trade associations of paper dealers and their members to adhere to price lists issued by the associations), enforcement denied in part and granted in part, 4 F.2d 457 (9th Cir. 1925), rev’d in part and FTC order enforced, 273 U.S. 52 (1927); Virginia Excelsior Mills, Inc., 54 F.T.C. 455 (1957) (condemning agreement of excelsior producers to establish common sales agent that set prices for all producers and allocated orders according to relative productive capacity of each producer), aff’d, 256 F.2d 538 (4th Cir. 1958); National Macaroni Manufac- turers Ass’n, 65 F.T.C. 583 (1964) (condemning agreement among pasta producers to fix the inputs used to make their products), aff’d, 345 F.2d 421 (7th Cir. 1965); American Medical Ass’n, 94 F.T.C. 701 (1979) (condemning AMA’s restrictions on truthful advertising and solicitation by its members), enforced as modified, 638 F.2d 443 (2d Cir. 1980), aff’d by an equally divided Court, 455 U.S. 676 (1982); Indiana Federation of Dentists, 101 F.T.C. 57 (1983) (condemning association’s efforts to prevent its members from complying with insurers’ requests for x-rays with insurance claims), enforcement denied and order vacated, 745 F.2d 1124 (7th Cir. 1984), rev’d and FTC order aff’d, 476 U.S. 447 (1986); Superior Court Trial Lawyers Ass’n, 107 F.T.C. 510 (1986) (condemning boycott designed to help effectuate agreement among attorneys to raise prices), enforcement denied and remanded, 856 F. 2d 226 (D.C. Cir. 1988), rev’d and FTC order aff’d, 493 U.S. 411 (1990); Massachusetts Board of Registration in Optometry, 110 F.T.C. 549, 604 (1988) (condemning restrictions on optometrists’ 2 No analytical exercise is more important to U.S. competition policy than defining the bounds of acceptable cooperation between direct rivals. Courts and commentators have written extensively on how Section 1 of the Sherman Act, 15 U.S.C. § 1, and Section 5 of the Federal Trade Commission Act (“FTC Act”), 15 U.S.C. § 45, apply to agreements involving competitors.1 The Federal Trade Commission (“the FTC” or “the Commission”) also has played a formative role in the evolution of horizontal restraints jurisprudence and policy.2 Our opinion in price and non-price advertising); Detroit Auto Dealers Ass’n, Inc., 111 F.T.C. 417 (1989) (condemning agreement among Detroit-area automobile dealers to close dealer showrooms on nights and weekends), aff’d in part and rev’d in part, 955 F.2d 457 (6th Cir.), cert. denied, 506 U.S. 973 (1992). In addition to developing doctrine through adjudication, the FTC has coauthored guidelines to help build the modern analytical framework for horizontal restraints. See Department of Justice and Federal Trade Commission, Antitrust Guidelines for Collaborations Among Competitors (Apr. 7, 2000) (“Collaboration Guidelines”); Department of Justice and Federal Trade Commission, Statements of Antitrust Enforcement Policy in Healthcare (Aug. 28, 1996); Department of Justice and Federal Trade Commis- sion, Antitrust Guidelines for the Licensing of Intellectual Property (Apr. 6, 1995). 3 On July 31, 2001, when the Commission announced the issuance of the complaint against Polygram, it also announced that it had accepted for public comment a consent agreement with Warner, settling similar allegations against Warner. On September 17, 2001, the Commission issued the final consent order against Warner, enjoining agreements with a competitor to fix prices or limit truthful, non-deceptive advertising or promotion for any audio or video product. Warner Communications Inc., Dkt. No. C-4025 (Sept. 17, 2001). 3 this matter provides our first adjudicative opportunity to revisit the issue of competitor collaboration since the Supreme Court’s decision in California Dental Ass’n v. Federal Trade Commission, 526 U.S. 756 (1999) (“CDA”), and the issuance of the Department of Justice and FTC Collaboration Guidelines. I. BACKGROUND The Commission issued its complaint in this matter on July 30, 2001. The complaint charges that the Respondents (hereinafter collectively referred to as “PolyGram”) engaged in unfair methods of competition in violation of Section 5 of the FTC Act by agreeing with competitor Warner Communications Inc. (“Warner”) to restrict price competition and forgo advertising.3 The complaint alleges that, after forming a joint venture (whose establishment the Commission does not challenge here) to collaborate in the distribution of audio and video recordings of a concert by the “Three Tenors” at the 1998 FIFA World Cup for soccer in Paris, PolyGram and Warner entered into a side agreement not to discount or advertise their previous Three Tenors products for a period of time 4 This opinion uses the following abbreviations: ID - Initial Decision of the Administrative Law Judge (“ALJ”). IDF - Numbered Findings of Fact in the ALJ’s Initial Decision CX - Complaint Counsel’s Exhibit RX - Respondents’ Exhibit JX - Joint Exhibits Tr. - Transcript of Trial before the ALJ We adopt the ALJ’s findings of fact to the extent such findings are not inconsistent with this opinion. 4 preceding and following the release of the new Three Tenors recording. The complaint alleges that these restrictions had the effect of restraining competition unreasonably, increasing prices, and injuring consumers. A. PolyGram PolyGram is a group of vertically integrated companies, affiliated with PolyGram N.V., engaged in the business of producing, marketing, and distributing recorded music and videos in the United States and worldwide. In 1998, PolyGram comprised Respondent PolyGram Holding, Inc. (“PolyGram Holding”); The Decca Record Company Limited (“Decca”) (now Respondent Decca Music Group Limited); PolyGram Records, Inc. (“PolyGram Records”) (predecessor of Respondent UMG Recordings, Inc. (“UMG”)); and PolyGram Group Distribution, Inc. (“PGD”) (predecessor of Respondent Universal Music & Video Distribution Corp. (“UMVD”)). IDF 7-11, 23.4 In December 1998, the Seagram Company Ltd. (“Seagram”) acquired PolyGram N.V. Seagram combined the music business of PolyGram N.V. (i.e., PolyGram) with its own music business to form Universal Music Group. Two years later, Seagram merged with Vivendi S.A. and Canal Plus S.A. to form Vivendi Universal S.A. (“Vivendi”). Each Respondent is now a subsidiary of Vivendi. IDF 6, 18. Decca is a music “label” that develops, acquires, and produces recorded music. In 1998, Decca was part of the PolyGram Classics & Jazz (“PolyGram Classics”) label group, a division of PolyGram Records. At all relevant times, 5 Since 1990, audio and video recordings of 3T1 have been distributed in the United States by PGD and its successor UMVD. PGD was responsible for deciding the wholesale price and advertising strategy for 3T1 in the United States. IDF 17. 5 Decca owned the copyright to the master recording of the first Three Tenors concert (“3T1”). IDF 14. PolyGram Classics was the PolyGram operating company responsible for United States sales of classical music produced by PolyGram. PolyGram Classics was responsible for marketing, promoting, pricing, and advertising 3T1 in the United States. IDF 12, 15. PGD provided the distribution and sales force for PolyGram Classics in the United States and executed PolyGram Classics’s marketing strategy at the retailer level. IDF 16. PolyGram Holding is the parent company of Respondents UMG and UMVD, and provides services to its subsidiaries, including legal, financial, business affairs, and human resources services. PolyGram Holding negotiated the collaboration between PolyGram and Warner with regard to the third Three Tenors World Cup concert (“3T3”). IDF 12-13. B. Warner Warner was PolyGram’s partner in the Three Tenors joint venture. Two Warner entities principally were involved in the conduct at issue here: Atlantic Recording Corp. (“Atlantic”), a Warner label that operates in the United States, and Warner Music International (“WMI”), which manages the music operations of Warner’s operating companies outside the United States. IDF 20-22. C. Factual Background The Three Tenors are world-renowned opera singers Jose Carreras, Placido Domingo, and Luciano Pavarotti. IDF 4-5. During the 1990s, the Three Tenors released three paired audio and video recordings derived from live concerts at the FIFA World Cup. PolyGram acquired the rights to distribute audio and video recordings of the first performance of the Three Tenors at the Baths of Caracalla in Rome in 1990.5 The trio’s first album became the best-selling classical record of 6 In 1994, 3T2 was the no. 2 and 3T1 was the no. 3 best-selling classical album (CX 587); in 1995, 3T2 was no. 1 and 3T1 was no. 5 (CX 588); in 1996, 3T2 was no. 4 and 3T1 was no. 5 (CX 589); and in 1997, 3T1 was no. 9 and 3T2 was no. 12 (CX 590). 7 Rudas is independent of PolyGram and Warner. See CX 380. 6 all time. IDF 27-29. In 1994, the Three Tenors performed a second World Cup concert at Dodger Stadium in Los Angeles. Warner acquired the rights to distribute audio and video recordings derived from the second concert (“3T2”). IDF 30, 32. In 1998, the Three Tenors performed a third World Cup concert in Paris. PolyGram and Warner entered into an agreement to collaborate in the distribution of the audio and video recordings of the third concert, with Warner distributing 3T3 in the United States and PolyGram distributing it in the rest of the world. IDF 59-60. Upon the release of 3T2 in 1994, and until 1998, PolyGram and Warner competed to sell their respective Three Tenors albums. IDF 34. In 1994, Warner launched an expensive and aggressive marketing campaign to support 3T2 in the United States and internationally. IDF 200-09. PolyGram responded to the release of 3T2 by promoting 3T1 aggressively in the United States and other markets, through advertising and price discounts. IDF 210-21. Sales of 3T1 audio and video products in the second half of 1994 increased over 250% compared with sales in the same period in 1993. JX 12. Despite the competition from 3T1, 3T2 was a business success for Warner. IDF 222. During 1996 and 1997, the Three Tenors held concerts in Tokyo, London, Munich, New York, Johannesburg, and Melbourne. PolyGram and Warner competed with each other throughout the world to capitalize on these concerts as an opportunity to drive sales of their Three Tenors products through various promotional activities. IDF 224-31. 3T1 and 3T2 were both among the best-selling classical recordings in the United States in 1994, 1995, 1996, and 1997. IDF 234.6 In 1996, PolyGram and Warner each began to negotiate separately with the concert promoter, Tibor Rudas (“Rudas”),7 for the rights to distribute the recordings of the next Three Tenors World Cup concert in 1998. PolyGram did not anticipate collaborating with Warner. IDF 54. Initially, Warner planned to distribute 3T3 without a collaboration with PolyGram: its Atlantic label proposed 8 Pavarotti was also under contract to record exclusively for Decca at the time of the 1994 3T2 concert. CX 224. In exchange for certain consideration, Decca agreed to waive its rights and allow Pavarotti to record for Warner. IDF 33. 9 PolyGram also sought to differentiate the 1998 concert by including a guest performer or original songs to be written by Andrew Lloyd Webber, Elton John, Stevie Wonder, or others, but these suggestions were rejected by the Three Tenors. IDF 75-76. 7 to distribute 3T3 in the United States, with WMI to distribute 3T3 in the rest of the world. IDF 52. The president of WMI, however, decided to pass on the project because he did not think that another Three Tenors album was a good investment. CX 366; Tr. 407-08. At that time, Pavarotti was under contract to record exclusively for PolyGram’s Decca label.8 In 1997, Warner asked Decca to release Pavarotti from his exclusive contract and permit him to record the 1998 World Cup concert for Warner. Instead, PolyGram proposed that Warner and PolyGram work together on the 3T3 project. Warner accepted this proposal. IDF 55-56. PolyGram and Warner were very concerned that the new Three Tenors album, scheduled for release in August 1998, would not be as original or commercially appealing as the 1990 and 1994 releases. IDF 73. They recognized that the commercial success of 3T3 would depend largely on having a repertoire that was distinct from that of the earlier Three Tenors recordings. IDF 66, 69. In their negotiations with Rudas, PolyGram and Warner sought the right to approve a significant part of the repertoire for the 1998 concert, but Rudas insisted that he and the artists should control the choice of songs. IDF 67-68.9 PolyGram and Warner ultimately agreed to forgo approval of the repertoire, and the contract with Rudas provided only that Rudas would consider “in good faith” their suggestions as to repertoire. IDF 68, 71-72. The collaboration between PolyGram and Warner took the following form: In a series of contracts dated October 14, 1997, in return for an $18 million advance and other consideration, Rudas licensed to Warner the worldwide audio, video, and home television rights to the 1998 concert. IDF 58. Then, in an 10 “Catalog” is a music industry term that refers to older albums that a record company continues to offer for sale. IDF 93. 8 agreement dated December 17, 1997, Warner licensed to PolyGram the rights to exploit 3T3 outside of the United States, with Warner (through its affiliate Atlantic) retaining the rights to exploit 3T3 within the United States. The contract provided that PolyGram would reimburse Warner for 50% of the $18 million advance paid to Rudas, and that Warner and PolyGram would share 50-50 the profits and losses from the 3T3 project. IDF 59-60. The contract also provided that Warner and PolyGram would have the right to market a Greatest Hits album and/or a Boxed Set incorporating the 1990, 1994, and 1998 Three Tenors recordings, but the joint venture agreement did not include the marketing rights to the existing 1990 and 1994 Three Tenors albums. JX-10-F; JX 11 at UMG001790 (in camera). The contract also contained a limited covenant not to compete, which stated that neither PolyGram nor Warner would release another Three Tenors recording for four years following the release of 3T3, unless such release was pursuant to this agreement. The contract expressly provided, however, that PolyGram and Warner each could continue to exploit its older Three Tenors products. IDF 62-63. Thus, the relationship of 3T1 and 3T2 to the joint venture was clear: ownership and marketing rights for both were outside the joint venture. The operating companies of both PolyGram and Warner began developing marketing campaigns for 3T1 and 3T2 in early 1998. They planned to capitalize on the upcoming Three Tenors concert and the new album as an opportunity to increase sales of their catalog Three Tenor products. IDF 102-05, 115-18.10 PolyGram and Warner grew concerned, however, that competition from the catalog Three Tenors recordings would reduce the sales of the new Three Tenors album. As a result, they feared that they would not recoup their $18 million investment. Tr. 485; JX 9-E; JX 94 at 94, 96; JX 100 at 72-73 (in camera); JX 102 at 43; CX 202. In March 1998, executives of PolyGram and Warner met and agreed to refrain from advertising or reducing prices of 3T1 or 3T2 audio or video products in all markets in the weeks surrounding the release of 3T3. They called this agreement the “moratorium” agreement. IDF 90-101, 107-13. Warner’s operating companies, however, continued with plans to launch a discounting campaign for 3T2 scheduled to run through December 1998. IDF 118. When PolyGram learned of this, it informed its operating companies that if Warner discounted 3T2, they were free to retaliate with price discounts on 3T1. IDF 119- 9 21, 128, 130. By June 1998, senior management at both PolyGram and Warner believed that the moratorium agreement was likely to fall apart. IDF 126-27, 129, 131-32. In June 1998, PolyGram and Warner also learned that – contrary to Rudas’s earlier statement that 3T3 would contain an all-new repertoire – the repertoire would substantially overlap with that of the older Three Tenors concerts. IDF 79- 81, 133. This unwelcome news added to PolyGram’s and Warner’s concerns that 3T3 would lose sales to 3T1 and 3T2 and would not be commercially successful. IDF 133-36. Later that month, PolyGram and Warner executives exchanged reassurances that the companies would forgo discounting and advertising of 3T1 and 3T2 during the launch of 3T3. IDF 137-44, 147. PolyGram and Warner subsequently issued written instructions to their operating companies worldwide that forbade price discounting and advertising of 3T1 and 3T2 from August 1, 1998 through October 15, 1998. IDF 148-53. In late July 1998, after the Paris concert but before the release of 3T3, the legal departments of PolyGram and Warner learned of the moratorium agreement. IDF 154. The establishment of the moratorium created evident discomfort for PolyGram’s attorneys, who raised concerns with PolyGram’s management about the moratorium’s legitimacy. CX 459; JX 94 at 170-79; RX 719 at 3-7. Shortly thereafter, PolyGram sent a letter to Warner purporting to disavow the existence of a moratorium; likewise, at the request of its counsel, Warner sent a letter to PolyGram purporting to reject the moratorium agreement. IDF 156-57, 160-63. These letters, however, were mere pretense, and the moratorium agreement remained in effect. IDF 158-59, 163-64. The companies complied with the moratorium. Between August 1, 1998 and October 15, 1998, neither PolyGram nor Warner reduced the prices of or funded advertising for its respective catalog Three Tenors products in the United States. IDF 170-76. The companies substantially complied with the moratorium outside the United States, as well. IDF 177-81. In the end, 3T3 was unsuccessful. Published reviews were generally unfavorable. IDF 167. Several music reviewers noted the overlap in repertoire between the 1998 Three Tenors album and the earlier Three Tenors recordings. IDF 166. Sales of 3T3 fell far short of the companies’ projections in 1997, when 10 they thought 3T3 would feature an all-new repertoire, and PolyGram and Warner lost millions of dollars on the project. Tr. 522-25. In 1999, Decca agreed to waive its exclusive rights to the recording services of Pavarotti to allow him to record a Three Tenors album for Sony. In October 1999, Sony released the album – which consisted of Christmas songs derived from a performance of the Three Tenors in Vienna – with no restriction on marketing activities by PolyGram or Warner in support of their catalog Three Tenors albums. IDF 196-99. D. The ALJ’s Initial Decision After pretrial discovery, ALJ James P. Timony conducted a one-week trial. Complaint Counsel called four live witnesses: Anthony O’Brien, from Atlantic; Rand Hoffman, from PolyGram Holding; Professor Catherine Moore, the director of the Music Business Program at New York University; and Dr. Stephen Stockum, an economist. Respondents called no live witnesses. Both parties introduced deposition testimony and numerous documents. The record closed on March 20, 2002. Following post-trial motions, Judge Timony issued an initial decision and a proposed order on June 20, 2002. Judge Timony’s decision ruled that the moratorium agreement constituted an unfair method of competition in violation of Section 5 of the FTC Act. The ALJ found that the moratorium agreement – created several months after the joint venture agreement between PolyGram and Warner – was not ancillary to the 3T3 joint venture because it was not an integral part of the joint venture or reasonably necessary to market the joint venture product. ID at 50-53. Instead, the ALJ found that the moratorium was a “naked agreement to fix prices and restrict output” that was properly subject to per se condemnation. ID at 54, 68. The ALJ also evaluated the moratorium under an abbreviated (or “quick look”) rule of reason analysis. He ruled that if the moratorium’s anticompetitive effects were “obvious,” the burden would shift to Respondents to show the procompetitive benefits of the restraint. ID at 54-55. Turning first to the agreement not to discount 3T1 and 3T2, the ALJ concluded that this arrangement constituted horizontal price fixing, which, as case law has recognized, “threatens 11 the efficient functioning of a market economy.” ID at 56. The ALJ found that PolyGram and Warner previously had competed by reducing the price of 3T1 and 3T2 – to the benefit of consumers – and that such an agreement to forgo discount- ing had “obvious anticompetitive potential.” ID at 56-57. The ALJ also concluded that the agreement to forgo advertising of 3T1 and 3T2 was presumptively anticompetitive. ID at 57. The ALJ explained that economic theory and empirical research showed that advertising restrictions result in higher prices to consumers, and that the evidence here showed that advertising was an important competitive tool used by PolyGram and Warner in marketing the Three Tenors products, creating additional demand and encouraging price discounting. ID at 57-58. The ALJ found that PolyGram and Warner intended that their advertising ban would conceal the better-value Three Tenors recordings so that consumers instead would purchase the higher-margin 3T3 release. Judge Timony concluded that the potential anticompetitive effect of this strategy was “obvious.” ID at 58. Turning next to Respondents’ efficiency justifications, the ALJ found that the Respondents failed to meet their burden of identifying legitimate procompeti- tive justifications. ID at 58-65, 68-69. He found that the parties’ principal motive for the moratorium was to shield 3T3 from competition to protect their profits, which he deemed to be an illegitimate justification. ID at 60. He also rejected Respondents’ other proffered justifications, finding that they were implausible and, even if plausible, were invalid because they were unsupported by the evidence in this case. ID at 61-65. Finally, the ALJ rejected Respondents’ contention that PolyGram withdrew from the moratorium and thus should not be held liable. ID at 65-66. The ALJ issued a cease and desist order enjoining Respondents for 20 years from again agreeing with a competitor to fix prices or to restrict advertising in connection with the sale of audio and video products, except under certain specified circumstances related to a joint venture. 12 E. Questions Raised by the Appeal Respondents appeal from the ALJ’s determination that their conduct violated Section 5 of the FTC Act. They also challenge the appropriateness of the ALJ’s cease and desist order. First, Respondents argue that the ALJ erred in concluding that the moratorium is illegal per se. They assert that the moratorium falls outside any well-established category of restraints subject to per se condem- nation. Rather, they contend, the Commission must analyze the moratorium under the rule of reason because the restrictions at issue were reasonably related to the purpose of a legitimate joint venture. Second, Respondents argue that, in applying the rule of reason, the ALJ erred by relying on a presumption of anticompetitive effects that shifts the burden to Respondents to show plausible procompetitive justifications. Respondents contend that the Supreme Court’s decision in CDA requires the FTC to offer proof of actual anticompetitive effect before the burden may be shifted to Respondents to justify the restraints. Third, Respondents argue that, even if the correct legal standard is that restraints categorized as “inherently suspect” warrant a presumption of anticom- petitive effects that shifts the burden to a defendant to show procompetitive justifications, the adoption of the moratorium in the context of a procompetitive joint venture dictates that the moratorium not be considered presumptively anticompetitive. Fourth, Respondents argue that their identification of “plausible” procompetitive justifications requires an assessment of the moratorium’s net competitive effects under a full rule of reason analysis. Fifth, Respondents argue that a cease and desist order is inappropriate here, because there is no basis for concluding that Respondents are likely to engage in similar conduct again. 11 The Commission’s authority under Section 5 of the FTC Act extends to conduct that violates the Sherman Act. See, e.g., Federal Trade Commission v. Motion Picture Advertising Serv. Co., 344 U.S. 392, 394-95 (1953); Fashion Originators’ Guild of America, Inc. v. Federal Trade Commission, 312 U.S. 457, 463-64 (1941). In the case at hand, our analysis under Section 5 is the same as it would be under Section 1 of the Sherman Act. 12 See State Oil Co. v. Khan, 522 U.S. 3, 20 (1997) (“State Oil”) (noting role of courts in antitrust law “in recognizing and adapting to changed circumstances and the lessons of accumulated experience”); Business Electronics Corp. v. Sharp Electronics Corp., 485 U.S. 717, 732 (1988) (use of term “restraint of trade” in Section 1 of Sherman Act “invokes the common law itself, and not merely the static content that the common law had assigned to the term in 1890”); 13 II. LEGAL FRAMEWORK Courts, enforcement agencies, and commentators long have strived to refine operational principles for applying the Sherman Act’s command that “[e]very contract, combination in the form of trust or otherwise, or conspiracy, in restraint of trade . . . is declared to be illegal.” 15 U.S.C. § 1. Jurisprudence, commentary, and enforcement experience concerning this prohibition provide the basic foundations for the Commission’s evaluation of horizontal restraints under Section 5 of the FTC Act.11 In this section we identify major aspects of the development of horizontal restraints doctrine and present the framework we will apply to the challenged restrictions in this matter. A. The Law of Horizontal Restraints The seemingly categorical language of Section 1 of the Sherman Act mentions none of the analytical concepts – “per se illegality,” “ancillarity,” “quick look,” or “full-blown rule of reason” – that appear in U.S. horizontal restraints jurisprudence. These concepts have evolved under the antitrust common law that Congress contemplated when it cast the nation’s antitrust commands in general terms and entrusted the federal courts and the FTC with developing the opera- tional content for these provisions. Over time, the courts and the FTC have refined that content to account for insights gained from adjudication experience and from developments in economic and legal learning.12 National Society of Professional Engineers v. United States, 435 U.S. 679, 688 (1978) (in adopting Sherman Act, Congress “expected the courts to give shape to the statute’s broad mandate by drawing on common-law tradition”). 13 In Standard Oil, the Court explained: [T]he standard of reason . . . was intended to be the measure used for the purpose of determining whether in a given case a particular act had or had not brought about the wrong against which [Sherman Act § 1] provided. 221 U.S. at 60. See also State Oil, 522 U.S. at 10 (“Although the Sherman Act, by its terms, prohibits every agreement in ‘restraint of trade,’ this Court has long recognized that Congress intended to outlaw only unreasonable restraints.”). 14 In Chicago Board of Trade, the Court said: The true test of legality is whether the restraint imposed is such as merely regulates and perhaps thereby promotes competition or whether it is such as may suppress or even destroy competition. 14 A number of tensions have marked the evolution of horizontal restraints doctrine and the pursuit of techniques for identifying restrictions that suppress competition. Perhaps most important, adjudicatory tribunals have struggled to attain an appropriate balance between achieving accuracy in individual cases, which generally requires fuller inquiry, and streamlining the law’s administration, which usually involves making simplifying assumptions and forgoing elaborate analysis when the conduct at issue ordinarily poses grave competitive dangers. In Standard Oil Co. v. United States, 221 U.S. 1 (1911), the Supreme Court made clear that Section 1 establishes a single, general principle governing trade restraints. The “rule of reason” is the touchstone for evaluating challenged conduct.13 As stated in Standard Oil and reiterated later in the same decade in Chicago Board of Trade v. United States, 246 U.S. 231 (1918), the purpose of courts in applying the rule of reason is to evaluate the impact of challenged behavior upon competition.14 246 U.S. at 238. 15 See State Oil, 522 U.S. at 21 (“[T]his Court has reconsidered its decisions construing the Sherman Act when the theoretical underpinnings of those decisions are called into serious question.”); see also Arizona v. Maricopa County Medical Society, 457 U.S. 332, 344 (1982) (“Once experience with a particular kind of restraint enables the Court to predict with confidence that the rule of reason will condemn it, it has applied a conclusive presumption that the restraint is unreasonable.”). 15 In articulating this principle, Standard Oil also endorsed a concept that earlier cases such as United States v. Trans-Missouri Freight Ass’n, 166 U.S. 290 (1897), and United States v. Addyston Pipe & Steel Co., 85 F. 271 (6th Cir. 1898), aff’d, 175 U.S. 211 (1899) (“Addyston Pipe”), had introduced and that retains vitality today: not all trade restraints require the same degree of fact-gathering and analysis. Standard Oil, 221 U.S. at 65. Within the general framework of the rule of reason, certain restraints might be recognized as being so inherently and commonly unreasonable that courts might dispense with an elaborate analysis and condemn them as illegal per se. See id. (noting that Trans-Missouri Freight and other precedent established that the “nature and character” of certain contracts create “a conclusive presumption” that the conduct violates the Sherman Act). Decisions about the appropriate form of inquiry would evolve over time as courts gained experience in evaluating specific business phenomena and accounted for commentary examining the rationale for and effects of various practices.15 Early decisions also yielded important analytical tools to help courts determine the appropriate form of inquiry for specific restraints. One of the most influential techniques appeared in Addyston Pipe in 1898. Seeking to avoid overinclusive application of Section 1, Judge (later Chief Justice) William Howard Taft introduced the concept of ancillarity. Addyston Pipe, 85 F. at 281-82. A simple (“naked”) agreement by rivals to set prices, allocate customers, or divide sales territories would be condemned summarily, but the adoption of a uniform pricing schedule as part of the operation of a partnership, which could provide services beyond the capability of any single individual, warranted more tolerant consideration because it was “ancillary” to a legitimate transaction. Even in times when enthusiasm for per se rules of liability grew, ancillarity played a crucial role in permitting firms to undertake efficient transactions without Sherman Act 16 For example, in Northern Pac. Ry. Co. v. United States, 356 U.S. 1 (1958) (“Northern Pacific”), the Supreme Court explained that “[t]his principle of per se unreasonableness . . . avoids the necessity for an incredibly complicated and prolonged economic investigation into the entire history of the industry involved, as well as related industries, in an effort to determine at large whether a particular restraint has been unreasonable – an inquiry so often wholly fruitless when undertaken.” Id. at 5. The idea that a conventional rule of reason inquiry entailed a vast analytical undertaking took root in the observation of Justice Brandeis in Chicago Board of Trade that a court in a rule of reason case must ordinarily consider the facts peculiar to the business to which the restraint is applied; its condition before and after the restraint was imposed; the nature of the restraint and its effect, actual or probable. The history of the restraint, the evil believed to exist, the reason for adopting the particular remedy, the purpose or end sought to be obtained, are all relevant facts. 246 U.S. at 238. This much-quoted formulation is often criticized as too comprehensive and open-ended to be helpful. See VII Areeda & Hovenkamp, Antitrust Law ¶ 1502, at 345. 16 condemnation. The willingness of contemporary horizontal restraints jurisprudence to consider efficiency rationales has descended substantially from this ancillarity principle. Following Chicago Board of Trade, particularly from the late 1930s through the early 1970s, the Supreme Court appeared to discern a sharp dichotomy between per se and reasonableness analysis – between summary condemnation (in which plaintiffs often prevailed if an agreement was proven) and an abyss of reasonableness analysis (from which defendants routinely emerged unscathed).16 The Court’s cases in this era reflected little sense that there were manageable alternatives between the poles. For a time, the acceptance of a dichotomy and the perceived absence of intermediate analytical approaches appear to have helped inspire the Court to categorize an ever wider array of conduct as per se illegal. By the early 1970s, the Court had found per se condemnation appropriate for a broad 17 In United States v. Socony-Vacuum Oil Co., 310 U.S. 150 (1940) (“Socony”), the Court endorsed a broad conception of horizontal collaboration that would be deemed to constitute per se illegal price-fixing. The Court said that “[u]nder the Sherman Act a combination formed for the purpose and with the effect of raising, depressing, fixing, pegging, or stabilizing the price of a commodity in interstate or foreign commerce is illegal per se.” Id. at 223. In a famous footnote, the Court explained that proof of actual anticompetitive effects was not necessary to establish illegality, noting that all price fixing arrangements are “banned because of their actual or potential threat to the central nervous system of the economy.” Id. at 224 & n. 59. 18 United States v. Topco Associates, Inc., 405 U.S. 596, 608-10 (1972); Timken Roller Bearing Co. v. United States, 341 U.S. 593, 597-98 (1951). 19 Klor’s, Inc. v. Broadway-Hale Stores, Inc., 359 U.S. 207, 212 (1959). 20 Albrecht v. Herald Co., 390 U.S. 145, 152-54 (1968) (maximum resale price maintenance); United States v. Arnold, Schwinn & Co., 388 U.S. 365, 379 (1967) (vertical territorial restrictions); Northern Pacific, 356 U.S. at 5-6 (tying). 21 See United States v. Joint Traffic Ass’n, 171 U.S. 505, 567-68 (1898) (Sherman Act not intended to proscribe all partnerships or the imposition of non- 17 range of horizontal arrangements affecting prices,17 the allocation of customers or territories,18 and various concerted refusals to deal.19 The Court’s treatment of vertical restraints exhibited similar trends.20 The inability to recognize intermediate approaches posed difficulties in an important category of cases. In some instances, restraints resembled conduct subject to summary condemnation but also appeared to promote the attainment of valuable efficiencies. While declining to surrender the administrability benefits of per se tests, courts searched for ways to distinguish unambiguously harmful restraints from conduct that arguably served legitimate ends. Even early Supreme Court decisions that endorsed a literalist reading of Section 1's ban on “every” contract in restraint of trade disavowed any aim to bar all agreements that in some sense limited the commercial freedom of the parties but also generated important efficiencies.21 As mentioned above, Addyston Pipe injected vital flexibility into competition covenants to facilitate the sale of good will in a business). 22 See discussion of Addyston Pipe at p. 15-16, supra. 23 The Court foreshadowed BMI in National Society of Professional Engineers v. United States, 435 U.S. 679 (1978) (“Professional Engineers”). In Professional Engineers the Court’s assessment of restraints contained in a profes- sional association’s code of ethics anticipated themes that BMI later emphasized. For example, the analysis in Professional Engineers resembles the characterization inquiry endorsed in BMI. The Court began by noting that the restriction in ques- tion “operates as an absolute ban on competitive bidding” and finding that “no elaborate industry analysis is required to demonstrate the anticompetitive character of such an agreement.” Id. at 692. The Court then considered the defendant’s “affirmative defense” that uninhibited competitive bidding “would lead to deceptively low bids, and would thereby tempt individual engineers to do inferior work with consequent risk to public safety and health.” Id. at 693. The Court rejected this defense, stating that the possibility that “competition is not entirely conducive to ethical behavior, . . . is not a reason, cognizable under the Sherman Act, for doing away with competition.” Id. at 696. 18 Section 1 analysis by introducing ancillarity as a means for sorting benign from pernicious restraints.22 In the mid- to late 1970s, the Court stepped back from the rigid categorical approach to Section 1 analysis that had prevailed since Socony. For horizontal restraints, the pivotal modern case was Broadcast Music, Inc. v. Columbia Broadcasting System, Inc., 441 U.S. 1 (1979) (“BMI”).23 Although the blanket copyright licenses challenged there were literally agreements to fix prices, the Court recognized that this fact alone did not establish that the practice was “price fixing” subject to the per se rule. Id. at 8-9. Rather, the Court acknowledged that before a court may condemn collaborative activity as per se illegal, it must conduct some assessment of whether the defendant had a legitimate business justification for the collaboration. The Court posed two central questions in attempting to characterize the activity: First, is the practice “‘plainly anti- competitive,’” id. at 8 (citation omitted), in that it “facially appears to be one that would always or almost always tend to restrict competition and decrease output”? Id. at 19-20. And, second, is the practice “designed to increase economic 24 Applying this analysis, the Court concluded that the blanket license was necessary to achieve the efficiencies of integration of sales, monitoring, and enforcement against unauthorized copyright use; thus, a “more discriminating” rule of reason analysis – rather than per se condemnation – was required. 441 U.S. at 20-24. 19 efficiency and render markets more, rather than less, competitive”? Id. at 20 (citation omitted).24 BMI abandoned the view that posits a sharp dichotomy between rule of reason and per se analysis and thus took a major step toward restoring unity to Section 1 analysis. BMI made explicit and transparent a characterization process that courts performed even during the dichotomy model’s apex. The dichotomy model placed all horizontal restraints in two boxes – one containing per se illegal acts and the other containing conduct that warranted a full reasonableness inquiry. To apply this framework in an individual case, the court had to make a threshold decision whether the arrangement at issue belonged in one box or the other. Unless the defendant conceded that its conduct fit exactly within a template of per se illegality established in earlier cases, the court was likely to confront arguments that the conduct could not be condemned summarily. To resolve such arguments, courts performed variants of the characterization exercise that BMI brought into full view. Under BMI and its progeny, however, characterization no longer necessarily determines the result of the case. Five years later, National Collegiate Athletic Ass’n v. Board of Regents of the University of Oklahoma, 468 U.S. 85 (1984) (“NCAA”), reinforced the teaching of BMI that courts must engage in an initial assessment of efficiency rationales before condemning conduct as per se illegal. In NCAA, the Court recognized that the agreements at issue there constituted horizontal price fixing and restrictions on output – categories of practices ordinarily condemned as per se illegal. Nonetheless, the Court declined to invoke the per se rule. Id. at 100-01. The Court noted that some horizontal restraints were “essential” to make the product (college football) available, id. at 101-02, and that a joint selling arrangement may have legitimate procompetitive efficiencies. Id. at 103 (citing BMI, 441 U.S. at 18-23). The Court held that, under these circumstances, a fair 25 See also Northwest Wholesale Stationers, Inc. v. Pacific Stationery & Printing Co., 472 U.S. 284, 295 (1985) (“Northwest Wholesale Stationers”) (Court declined to apply per se rule to group boycott by a wholesale purchasing cooperative that expelled one of its members, noting that “such cooperative arrangements would seem to be ‘designed to increase economic efficiency and render markets more, rather than less, competitive’” because “[t]he arrangement permits the participating retailers to achieve economies of scale . . ., and also ensures ready access to a stock of goods that might otherwise be unavailable on short notice”) (quoting BMI, 441 U.S. at 20). 26 When direct evidence of actual effects can be shown, elaborate market definition is unnecessary. Federal Trade Commission v. Indiana Federation of Dentists, 476 U.S. 447, 460-61 (1986); Todd v. Exxon Corp., 275 F.3d 191, 206 (2d Cir. 2001) (“an actual adverse effect on competition . . . arguably is more direct evidence of market power than calculations of elusive market share figures”); Re/Max International, Inc. v. Realty One, Inc., 173 F.3d 995, 1018 (6th Cir. 1999) (“an antitrust plaintiff is not required to rely on indirect evidence of a defendant’s monopoly power, such as high market share within a defined market, when there is direct evidence that the defendant has actually set prices or excluded competition”). 27 See William J. Kolasky, Jr., Counterpoint: The Department of Justice’s “Stepwise” Approach Imposes Too Heavy a Burden on Parties to Horizontal Agreements, 12 Antitrust 41, 44-45 (Spring 1998) (the “quick look” approach “is simply an application of the standard rule of reason analysis in circumstances where the effect on competition is apparent and the defendant’s procompetitive explanation for it is facially unconvincing, thus allowing the court 20 evaluation of the competitive character of the restraints at issue required consideration of the NCAA’s claimed justifications. Id.25 NCAA also established that, even if summary condemnation under the per se rule is inappropriate, full rule of reason analysis is not necessarily the alternative. Full rule of reason analysis often entails defining the market and examining market power, inquiries that usually require elaborate analysis.26 Sometimes a restraint’s competitive harm is evident after an abbreviated rule of reason analysis, obviating elaborate proof under the full rule of reason.27 In NCAA, for example, to end, i.e., truncate, its analysis”). 21 the Court held that “when there is an agreement not to compete in terms of price or output, ‘no elaborate industry analysis is required to demonstrate the anticom- petitive character of such an agreement.’” Id. at 109 (quoting Professional Engineers, 435 U.S. at 692). Although the Court in NCAA went on to consider asserted efficiencies of the association’s restrictions, id. at 113-17, it did so within the framework of a truncated analysis, without need for a full rule of reason approach. The Court first noted that there was no reason to believe that the restrictions on the product in question (i.e., television rights to college football games) could bring efficiencies to the sale of that product. Id. at 113-15. Next, the Court rejected out of hand arguments that restrictions on one product (television rights) could be justified by the prospect of enhancing sales of another product (live attendance tickets). Id. at 115-17. While noting that this argument, too, lacked a factual underpinning, the Court held that the “more fundamental reason” for rejecting such an argument is that it is “inconsistent with the basic policy of the Sherman Act” to insulate a product from competition in this manner. Id. at 116-17. In other words, such an argument is not cognizable as a matter of law. The NCAA Court also made clear that a proffered justification for an otherwise unlawful restraint must be reasonably “tailored” to serve the asserted procompetitive interests. In rejecting the NCAA’s arguments that the challenged restrictions could help to preserve competitive balance among amateur teams, the Court emphasized that a variety of less restrictive alternatives were available that would have served that goal at least as well. Id. at 119. See also Collaboration Guidelines, supra note 2, at § 3.36(b) (“[I]f the participants could have achieved or could achieve similar efficiencies by practical, significantly less restrictive means, then the Agencies conclude that the relevant agreement is not reasonably necessary to their achievement.”); XI Herbert Hovenkamp, Antitrust Law ¶ 1913 (1998). Similarly, in Federal Trade Commission v. Indiana Federation of Dentists, 476 U.S. 447 (1986) (“IFD”), the Court did not require extensive market analysis to ascertain the competitive harm resulting from practices that it considered obviously anticompetitive, but instead focused on whether there was an efficiency 22 justification for such practices. There, the Court found that “‘no elaborate industry analysis is required to demonstrate the anticompetitive nature of’” an agreement among dentists to withhold from their customers a desired service (providing x- rays to insurers in conjunction with insurance claim forms); accordingly, “[a]bsent some countervailing procompetitive virtue – such as, for example, the creation of efficiencies in the operation of a market or the provision of goods and services, . . . – such an agreement limiting consumer choice by impeding the ‘ordinary give and take of the market place,’ . . . cannot be sustained under the Rule of Reason.” Id. at 459 (quoting Professional Engineers, 435 U.S. at 692). Turning to IFD’s justification – that allowing insurance companies to make coverage decisions on the basis of x-rays would harm the quality of care provided to patients – the Court found this argument legally and factually flawed: The argument is, in essence, that an unrestrained market in which consumers are given access to the information they believe to be relevant to their choices will lead them to make unwise or even dangerous choices. Such an argument amounts to ‘nothing less than a frontal assault on the basic policy of the Sherman Act.’ [Professional Engineers, 435 U.S. at 695.] Moreover, there is no particular reason to believe that the provision of information will be more harmful to consumers in the market for dental services than in other markets. 476 U.S. at 463. Because IFD’s justification did not withstand scrutiny, the Court concluded that the challenged practice was unlawful. Id. at 465-66. BMI, NCAA, and IFD indicated that the evaluation of horizontal restraints takes place along an analytical continuum in which a challenged practice is examined in the detail necessary to understand its competitive effect. Neverthe- less, these cases did not provide a clear structure for the required analysis. In 1988, the Commission itself sought to provide a structured framework in Massachusetts Board of Registration in Optometry, 110 F.T.C. 549 (1988) (“Mass. Board”): First, we ask whether the restraint is “inherently suspect.” In other words, is the practice the kind that appears likely, absent an efficiency 28 110 F.T.C. at 604 (emphasis in original). The Commission applied the Mass. Board framework the following year in Detroit Auto Dealers Ass’n, 111 F.T.C. 417, 492-501 (1989), and ruled that an agreement among Detroit automobile dealers to close dealer showrooms on nights and weekends unreasonably restrained trade. The Sixth Circuit rejected the Commission’s conclusion that the restraint was “inherently suspect” as an improper application of the per se rule. Detroit Auto Dealers Ass’n, Inc. v. Federal Trade Commission, 955 F.2d 457, 470-71 (6th Cir. 1992). In particular, the court criticized the Commission’s reliance on Robert Bork’s argument (in his treatise, The Antitrust Paradox (1978)) that there is no economic difference between an agreement to limit shopping hours and an agreement to increase price. 955 F.2d at 470. The Commission’s analysis, however, rested upon more than citations to Judge Bork’s book. The Commission found ample record evidence demonstrating that showroom hours are an important basis on which dealers compete for customers. For example, it was undisputed that Detroit was the only metropolitan area in the country in which almost all dealers were closed on weekends. 111 F.T.C. at 497- 98. Although it disagreed with the Commission’s “inherently suspect” categorization, the court upheld the Commission’s ruling that the limitation of showroom hours was an unreasonable restraint of trade, because hours of 23 justification, to “restrict competition and decrease output”? . . . If the restraint is not inherently suspect, then the traditional rule of reason, with attendant issues of market definition and power, must be employed. But if it is inherently suspect, we must pose a second question: Is there a plausible efficiency justification for the practice? That is, does the practice seem capable of creating or enhancing competition (e.g., by reducing the costs of producing or marketing the product, creating a new product, or improving the operation of the market)? Such an efficiency defense is plausible if it cannot be rejected without extensive factual inquiry. If it is not plausible, then the restraint can be quickly condemned. But if the efficiency justifi- cation is plausible, further inquiry – a third inquiry – is needed to determine whether the justification is really valid. If it is, it must be assessed under the full balancing test of the rule of reason. But if the justification is, on examination, not valid, then the practice is unrea- sonable and unlawful under the rule of reason without further inquiry – there are no likely benefits to offset the threat to competition.28 operation are a basis of competition among automobile dealers, and because respondents failed to advance valid justifications for the restraint. 955 F. 2d at 471-72. 29 These cases are better understood as being consistent with the view of Sherman Act Section 1 analysis articulated in NCAA, IFD, and Mass. Board – that the court must consider proffered efficiencies before condemning a particular restraint. In SCTLA, the Court considered and rejected claimed efficiencies and other justifications before concluding that the challenged conduct (a boycott to force an increase in the compensation of court-appointed counsel) was a naked restraint on price and output falling within the per se category. 493 U.S. at 423- 24. In Palmer, the Court held that an agreement between competitors to divide markets and share revenues was per se illegal. The Palmer defendants did not argue that the agreement yielded procompetitive efficiencies or a new product. Instead, they tried to avoid liability by contending that the traditional ban on horizontal agreements to allocate sales territories did not apply if a firm agreed with a rival not to enter a market that it previously had not served. In such circumstances, the Supreme Court had little difficulty condemning the agreement outright. 498 U.S. at 49-50. 30 Alternatively, the Commission found the restraints on price advertising illegal under an abbreviated rule of reason analysis. The Commission also found the association’s restraints on non-price advertising illegal under an abbreviated rule of reason analysis. 121 F.T.C. at 320-21. 24 See also VII Areeda & Hovenkamp, Antitrust Law ¶ 1511c. The Commission later retreated from the Mass. Board approach in Cali- fornia Dental Ass’n, 121 F.T.C. 190 (1996), applying a per se rule and the rule of reason as “separate categories” of analysis. Id. at 299. The Commission reasoned that the Supreme Court had returned to such a categorical approach in two 1990 cases finding per se violations, Federal Trade Commission v. Superior Court Trial Lawyers Ass’n, 493 U.S. 411 (1990) (“SCTLA”), and Palmer v. BRG of Georgia, Inc., 498 U.S. 46 (1990) (“Palmer”).29 121 F.T.C. at 299. Thus, the Commission held that the dental association’s ethical rules restricting price advertising (which precluded, e.g., advertising that characterized a dentist’s fees as low or reasonable) were per se illegal. Id. at 307.30 The Ninth Circuit disagreed with the 31 CDA was the first case since BMI in which the Court found that the evidence was insufficient to condemn a basic horizontal restraint. In the eleven years following BMI, the Court issued six consecutive opinions finding the evidence sufficient to condemn the restraint. See Catalano, Inc. v. Target Sales, Inc., 446 U.S. 643 (1980) (per curiam) (“Catalano”); Maricopa, 457 U.S. 332; NCAA, 468 U.S. 85; IFD, 476 U.S. 447; SCTLA, 493 U.S. 411; Palmer, 498 U.S. 46. In each of these cases, except for NCAA, the Court reversed the court of appeals. But see Northwest Wholesale Stationers, 472 U.S. 284 (Court reversed circuit court’s ruling that wholesale cooperative’s expulsion of member warranted condemnation as per se illegal group boycott). 25 Commission’s per se approach and held that the advertising restrictions were properly condemned under an abbreviated rule of reason analysis, because they were facially anticompetitive and because CDA’s purported procompetitive justifications, although plausible, lacked evidentiary support. California Dental Ass’n v. Federal Trade Commission, 128 F.3d 720 (9th Cir. 1997). While the Supreme Court rejected the Ninth Circuit’s analysis as too abbreviated, the Court’s opinion leaves no doubt that it views Section 1 analysis as a continuum, rather than a series of distinct boxes (per se, quick look, full rule of reason). California Dental Ass’n v. Federal Trade Commission, 526 U.S. 756 (1999).31 In CDA, the Court explicitly acknowledged, for the first time, that its prior cases support an abbreviated or “quick look” rule of reason analysis. Id. at 770- 71. The Court recognized that advertising restrictions normally harm competition and consumers, but noted that CDA had advanced a number of reasons why its restrictions might nonetheless have served procompetitive purposes in light of the circumstances and context. Id. at 773. The restrictions did not ban advertising completely, id., and were designed on their face to avoid false or deceptive advertising and therefore “might plausibly be thought to have a net procompetitive effect, or possibly no effect at all on competition.” Id. at 771. Thus, the Court found that the anticompetitive effect of the restrictions on professional advertising was not obvious. Id. at 771, 778. The Court emphasized the professional context of the case before it, questioning whether market forces “normally” found in the commercial world apply to professional advertising, especially given that the market at issue was “characterized by striking disparities between the information 32 The majority opinion used the word “professional” more than 20 times. Respondents’ attempt to downplay the professional setting of CDA ignores this striking fact. 33 Although the Court criticized the Ninth Circuit for prematurely shifting the evidentiary burden to CDA to “adduce hard evidence of the pro- competitive nature of its policy,” 526 U.S. at 776, the Supreme Court’s own discussion repeatedly reflects the premise that CDA had identified potential justifications that not only were plausible in theory but also had some grounding in actual experience. See id. at 771 (“The restrictions on both discount and nondiscount advertising are, at least on their face, designed to avoid false or deceptive advertising in a market characterized by striking disparities between the information available to the professional and the patient.”); id. at 772 (“In a market for professional services, in which advertising is relatively rare and the comparability of service packages not easily established, the difficulty for customers or potential competitors to get and verify information about the price and availability of services magnifies the dangers to competition associated with misleading advertising.”); id. at 773 (“The existence of such significant challenges to informed decision making by the customer for professional services immediate- ly suggests that advertising restrictions arguably protecting patients from mislead- ing or irrelevant advertising call for more than cursory treatment as obviously comparable to classic horizontal agreements to limit output or price competi- tion.”); id. at 773-74 (“[T]he particular restrictions on professional advertising could have different effects from those ‘normally’ found in the commercial world, even to the point of promoting competition by reducing the occurrence of unverifiable and misleading across-the-board discount advertising.”); id. at 774 (“[T]he discipline of specific examples may well be a necessary condition of plausibility for professional claims that for all practical purposes defy comparison shopping.”); id. at 775 (“It might be, too, that across-the-board discount advertise- ments would continue to attract business indefinitely, but might work precisely because they were misleading customers . . . .”); id. at 778 (the Ninth Circuit 26 available to the professional and the patient.” Id. at 771-74.32 The Court con- cluded that, under these circumstances, and in the absence of any empirical evidence supporting the theoretical basis for a presumption of anticompetitive effects, CDA’s identification of plausible procompetitive justifications precluded the “indulgently abbreviated” review of the Ninth Circuit. Id. at 774-78.33 “failed to explain why it gave no weight to the countervailing, and at least equally plausible, suggestion that restricting difficult-to-verify claims about quality or patient comfort would have a procompetitive effect by preventing misleading or false claims that distort the market”). 34 On remand before the Ninth Circuit, the Commission argued that citations in the CDA record to a small fraction of the economic evidence relevant to the effects of the advertising restrictions provided an adequate basis to condemn the restraints at issue, and alternatively sought a remand to the FTC to develop a fuller record. The Ninth Circuit concluded that such evidence was not adequate to establish the likelihood of anticompetitive effects in this context, and declined to allow the Commission a “second bite at the apple” by remanding. California Dental Ass’n v. Federal Trade Commission, 224 F.3d 942, 950-52, 958 (9th Cir. 2000). In contrast to CDA, the record in the instant case contains a full discussion of the relevant economic literature. See infra note 52 and accompanying text. 27 The Court remanded for a more extended examination of the “tendency of these professional advertising restrictions.” Id. at 781. The Court specified that this did not necessarily call for the fullest market analysis. Id. at 780. “The truth,” said the Court, “is that our categories of analysis of anticompetitive effect are less fixed than terms like ‘per se,’ ‘quick look,’ and ‘rule of reason’ tend to make them appear.” Id. at 779. Rather, the Court indicated that rule of reason analysis should be flexible: As the circumstances here demonstrate, there is generally no categorical line to be drawn between restraints that give rise to an intuitively obvious inference of anticompetitive effect and those that call for more detailed treatment. What is required, rather, is an enquiry meet for the case, looking to the circumstances, details, and logic of a restraint. Id. at 780-81.34 CDA stops short of providing a complete analytical framework for the rule of reason inquiry, but gives important guidance about how abbreviated rule of reason analysis is to be conducted. Notably, CDA does not require a showing of 35 The Court focused on the restraint itself, identifying “the likelihood of anticompetitive effects” as that which must be examined under an abbreviated rule of reason analysis, 526 U.S. at 771 (emphasis added), and thus belied any claim that a showing of actual anticompetitive effect is required. Before each tribunal in CDA, including the Supreme Court, the dentists had argued that their restraints could not be condemned without proof that the dentists exercised power in appropriately defined markets. See, e.g., Brief of Petitioner California Dental Association, 28, 42-43 (Nov. 10, 1998). The Supreme Court’s CDA opinion contains no hint that the error below was failure to conduct a plenary analysis of market power. Indeed, the Court’s description of quick look analysis as that by which “an observer with even a rudimentary understanding of economics could conclude that the arrangements in question would have an anticompetitive effect on customers and markets,” 526 U.S. at 770, reveals that proof of market power is not a necessary element of this analysis. See also Stephen Calkins, California Dental Association: Not a Quick Look But Not the Full Monty, 67 Antitrust L.J. 495, 496 (2000) (“The most important lesson of CDA is that the defendant’s principal argument throughout the proceeding – that the Commission could prohibit its restraints only through elaborate, formal proof of market power – was rejected.”). 36 Because the Court relied on literature concerning professions other than dentistry, 526 U.S. at 771-73, the Court presumably would allow evidence concerning analogous professional markets. 28 actual anticompetitive effects or proof of market power.35 Its principal lesson is that rule of reason analysis must be sensitive to context and distinct characteristics of particular markets, particularly those involving professional services, in evaluating whether general rules of economic theory can be expected to apply. When, as in that case, the defendant articulates plausible reasons why its restrictions may not result in competitive harm and may result in cognizable procompetitive benefits, then the plaintiff’s showing of likely anticompetitive effects should have an “empirical” foundation, whether based on evidence specific to a particular case or in empirical studies of similar markets. Id. at 775 n. 12.36 Even in such cases, however, the plaintiff need not necessarily address the full range of issues regarding market conditions, if an “enquiry meet for the case” permits “a confident conclusion about the principal tendency of a restriction.” Id. at 781. CDA does not preclude – indeed, it is consistent with – the Commission’s 37 This synthesis addresses the analytical steps when the plaintiff seeks to avoid pleading and proving market power. It does not address the analysis when market power is at issue. 29 approach in Mass. Board, and it provides guidance about how that approach should be pursued. B. Synthesis As embodied most recently in CDA and in our Collaboration Guidelines, the development of modern horizontal restraints jurisprudence suggests an analytic framework that proceeds by several identifiable steps. These steps reflect the general principle that antitrust law proscribes only conduct that is likely to harm consumers. In most cases, conduct cannot be adjudged illegal without an analysis of its market context to determine whether those engaged in the conduct or restraint are likely to have sufficient power to harm consumers. In a smaller but significant category of cases, scrutiny of the restraint itself is sufficient to find liability without consideration of market power.37 A plaintiff may avoid full rule of reason analysis, including the pleading and proof of market power, if it demonstrates that the conduct at issue is inherently suspect owing to its likely tendency to suppress competition. Such conduct ordinarily encompasses behavior that past judicial experience and current economic learning have shown to warrant summary condemnation. If the plaintiff makes such an initial showing, and the defendant makes no effort to advance any competitive justification for its practices, then the case is at an end and the practices are condemned. If the challenged restrictions are of a sort that generally pose significant competitive hazards and thus can be called inherently suspect, then the defendant can avoid summary condemnation only by advancing a legitimate justification for those practices. Such justifications may consist of plausible reasons why practices that are competitively suspect as a general matter may not be expected to have adverse consequences in the context of the particular market in question; or they may consist of reasons why the practices are likely to have beneficial effects for consumers. 38 Although it has earlier roots, the concept of cognizability as a principle limiting the types of justifications has been clearly articulated at least since Professional Engineers, where the Supreme Court endorsed the view that certain types of defenses or justifications did not warrant consideration: We are faced with a contention that a total ban on competitive bidding is necessary because otherwise engineers will be tempted to submit deceptively low bids. Certainly, the problem of professional deception is a proper subject of an ethical canon. But, once again, the equation of competition with deception, like the similar equation with safety hazards, is simply too broad; we may assume that competition is not entirely conducive to ethical behavior, but that is not a reason, cognizable under the Sherman Act, for doing away with competition. 435 U.S. at 696. See also IFD, 476 U.S. at 463 (citing Professional Engineers in rejecting claim that competition would lead to “dangerous choices” because “there is no particular reason to believe” that consumers cannot digest the information competition provides); Collaboration Guidelines, supra note 2, at § 3.2 (“Some claims – such as those premised on the notion that competition itself is unreason- able – are insufficient as a matter of law . . . .”); compare Thomas G. Krattenmaker, Per Se Violations in Antitrust Law: Confusing Offenses With Defenses, 77 Geo. L.J. 165 (1988) (cases considered to identify “per se” offenses in antitrust analysis are best interpreted as identifying defenses that cannot redeem challenged behavior). 30 At this early stage of the analysis, the defendant need only articulate a legitimate justification. See CDA, 526 U.S. at 775 & n. 12. While the defendant at this point is not obligated to prove competitive benefits, id., the proffered justifications must be both cognizable under the antitrust laws and at least facially plausible. The first element, cognizability, allows the deciding tribunal to reject proffered justifications that, as a matter of law, are incompatible with the goal of antitrust law to further competition.38 Cognizable justifications ordinarily explain how specific restrictions enable the defendants to increase output or improve product quality, service, or innovation. By contrast, courts since the earliest decades of the Sherman Act have identified classes of justifications that, because they contradict the procompetition aims of the antitrust laws, will not save restraints from condemnation. For example, a defendant cannot defend restraints 39 See, e.g., Socony, 310 U.S. at 224 & n. 59 (“Whatever economic justification particular price-fixing agreements may be thought to have, the law does not permit an inquiry into their reasonableness.”); United States v. Trenton Potteries Co., 273 U.S. 392, 397-98 (1927) (“The reasonable price fixed today may through economic and business changes become the unreasonable price of tomorrow. . . . [I]n the absence of express legislation requiring it, we should hesitate to adopt a construction making the difference between legal and illegal conduct in the field of business relations depend on so uncertain a test as whether prices are reasonable . . . .”). 40 See IFD, 467 U.S. at 463-64 (confirming that, even in markets for professional services such as dentistry and engineering, there is no reason to believe that informed consumers will make unwise tradeoffs between quality and price); Professional Engineers, 435 U.S. at 696 (“[T]he Rule of Reason does not support a defense based on the assumption that competition itself is unreasonable.”). 41 See Catalano, 446 U.S. at 649 (refusing to recognize defense based on argument that limits on credit terms would promote new entry by raising price of product). 31 of trade on the ground that the prices the conspirators set were reasonable,39 that competition itself is unreasonable or leads to socially undesirable results,40 or that price increases resulting from a trade restraint would attract new entry.41 Of particular relevance here, the Supreme Court has recognized that a defendant cannot justify curbing access to a more-desired product to induce consumers to purchase larger amounts of a less-desired product. See NCAA, 468 U.S. at 116-17. Such justifications are not cognizable and require no further analysis. The second necessary element of legitimacy is plausibility. To be legitimate, a justification must plausibly create or improve competition. A justification is plausible if it cannot be rejected without extensive factual inquiry. The defendant, however, must do more than merely assert that its purported justification benefits consumers. Although the defendant need not produce detailed evidence at this stage, it must articulate the specific link between the challenged restraint and the purported justification to merit a more searching 42 As a practical matter, many of the claimed efficiencies likely will involve claims of “ancillarity.” See supra note 22 and accompanying text, supra Part II. A (describing development of ancillarity concept in antitrust analysis as tool for identifying restraints that increase efficiency). Although post-BMI cases generally speak of “efficiency,” the ancillary restraints doctrine retains its vitality in evaluating efficiency claims. The concept of ancillarity is implicit in our Collaboration Guidelines, see supra note 2, which recognize that restraints that otherwise might be considered illegal per se warrant more elaborate analysis when they are reasonably related to, and reasonably necessary for the achievement of, procompetitive benefits. Collaboration Guidelines, at § 1.2. Moreover, whether or not expressed in terms of ancillarity, the link between defendant’s “plausible” justification and a cognizable benefit must be clear. Unless it leads to a cognizable benefit, a proffered justification is irrelevant to the analysis. 43 Although this stage and the preceding inquiry could be combined, we think it analytically superior and consistent with the relevant case law to first screen the purported justification for legitimacy before engaging in a more extensive, and therefore longer and more resource-intensive, inquiry whether detailed analysis supports or refutes the justification. Antitrust courts have long held that preliminary analysis of purported justifications is appropriate. See, e.g., supra Part II.A. (discussing NCAA and IFD) and notes 38-39 and accompanying text (citing relevant cases). 32 inquiry into whether the restraint may advance procompetitive goals, even though it facially appears of the type likely to suppress competition.42 When the defendant advances such cognizable and plausible justifications, the plaintiff must make a more detailed showing that the restraints at issue are indeed likely, in the particular context, to harm competition.43 Such a showing still need not prove actual anticompetitive effects or entail “the fullest market analysis.” CDA, 526 U.S. at 779. Depending upon the circumstances of the cases and the degree to which antitrust tribunals have experience with restraints in particular markets, such a showing may or may not require evidence about the particular market at issue, but at a minimum must entail the identification of the theoretical basis for the alleged anticompetitive effects and a showing that the effects are indeed likely to be anticompetitive. See id. at 775 n.12. Such a showing may, for example, be based on a more detailed analysis of economic 44 In CDA, the partial restraints on professional advertising at issue could not be condemned without more evidence than the FTC provided. Accord- ing to the Supreme Court, the court of appeals failed to test the dentists’ proposed justification to determine whether the restraints themselves had “a net procompeti- tive effect, or possibly no effect at all on competition.” 526 U.S. at 771. In terms of the synthesis outlined here, the dentists prevailed either because (1) it was incorrect, without more evidence, to assume that restraints inherently suspect in “normal” (id. at 773) markets were similarly suspect in a professional setting, or (2) the restraints at issue had a plausible and cognizable justification that, given the complex nature of professional advertising, could not be rebutted by assump- tion alone. In either case, the burden on the Commission was the same: it was required to show why the presumption of likely anticompetitive effects that applies in non-professional markets also applied in the professional setting of CDA. 45 See, e.g., IFD, 476 U.S. at 459 (once plaintiff has met burden of showing likely anticompetitive effects, defendant must show “countervailing procompetitive virtue”); Law v. National Collegiate Athletic Assn, 134 F.3d 1010, 1019 (10th Cir.) (“Law”) (discussing shifting burdens of proof in rule of reason 33 learning about the likely competitive effects of a particular restraint, in markets with characteristics comparable to the one at issue. The plaintiff may also show that the proffered procompetitive effects could be achieved through means less restrictive of competition. The defendant, of course, can introduce evidence to refute the plaintiff’s arguments or to show that detailed evidence supports its proffered justification. Applying a flexible analysis “meet for the case,” the tribunal at this stage must ascertain whether it can draw “a confident conclusion about the principal tendency of a restriction” regarding competition. Id. at 781.44 The plaintiff has the burden of persuasion overall, but not necessarily the burden with respect to each step of this analysis. If the plaintiff satisfies its initial burden of showing that the practices in question are inherently suspect, then the defendant must come forward with a substantial reason why there are offsetting procompetitive benefits. If the defendant articulates a legitimate (i.e., cognizable and plausible) justification, then the plaintiff must address the justification, and provide the tribunal with sufficient evidence to show that anticompetitive effects are in fact likely, before the evidentiary burden shifts to the defendant.45 At this cases), cert. denied, 525 U.S. 822 (1998). 46 Cf. United States v. Baker Hughes Inc., 908 F.2d 981, 991 (D.C. Cir. 1990) (applying this principle in merger case). 47 The DOJ/FTC Collaboration Guidelines, supra note 2, draw upon the case law discussed above in providing an analytical structure for evaluating joint venture activity. The Agencies’ analysis “begins with an examination of the nature of the relevant agreement.” Collaboration Guidelines, at § 1.2. First, the Agencies ask whether the agreement is potentially per se illegal – i.e., is “of a type that always or almost always tends to raise price or reduce output.” Id. at § 3.2. If the answer is yes, then the Agencies consider proffered justifications. An agreement will escape per se challenge if it “is reasonably related to [efficiency- enhancing] integration and reasonably necessary to achieve its procompetitive benefits.” Id. The Collaboration Guidelines explain that before accepting proffered justifications, the Agencies undertake a limited factual inquiry to determine whether claimed justifications that are plausible in theory are plausible in the context of a particular collaboration, and that “[s]ome claims – such as those premised on the notion that competition itself is unreasonable – are insufficient as a matter of law.” Id. Following CDA, the Collaboration Guidelines specify that rule of reason analysis “entails a flexible inquiry and varies in focus and detail depending on the nature of the agreement and market circumstances.” Id. at § 3.3 (citations omitted). The Collaboration Guidelines also recognize that full rule of reason analysis may not be required: “[W]here the likelihood of anticompetitive harm is evident from the nature of the agreement, . . . then, absent overriding benefits that 34 stage, the defendant’s burden to respond will likely depend in individual cases upon the quality and amount of evidence that the plaintiff has produced to illuminate the competitive dangers of the defendant’s conduct.46 The defendant also has the burden of producing factual evidence in support of its contentions, including documents within its control. The existence of a joint venture or other collaboration is simply one circumstance to be considered in assessing the competitive effects of a challenged restraint.47 If a joint venture results in competitive benefits, such as the could offset the competitive harm, the Agencies challenge such agreements without a detailed market analysis.” Id. (citations omitted). The Collaboration Guidelines indicate that the underlying issue is the extent to which a challenged restraint in fact likely assists the parties in achieving efficiencies in the market circumstances at issue. Id. at § 3.36. 35 introduction of innovative products or the achievement of production efficiencies, then such benefits are a proper part of the antitrust analysis. But proffered justifications still must be analyzed under the framework stated above, and will entitle the defendant to a fuller review only if they are cognizable and are factually supported to the degree necessary in light of the plaintiff’s demonstration of likely anticompetitive effects. Our intended contribution in this synthesis is to specify more fully the analytical principles that we perceive to be embedded in the case law and our own guidelines and to refine the methodology for applying those principles in practice. Our synthesis thus responds to the need in modern competition policy to devise analytical tests that are sound in substance, transparent in revealing their operational criteria, and administrable in the routine analysis of antitrust disputes. III. ANALYSIS OF THE CHALLENGED RESTRAINTS Respondents argue that because the moratorium was “ancillary to a procom- petitive joint venture, that agreement cannot be deemed ‘presumptively anticompe- titive,’” Respondents’ Opening Brief at 41, and their practices cannot be held illegal without evidence of actual anticompetitive effect. Id. at 32. Respondents also argue that their identification of plausible procompetitive justifications means that their practices cannot be held illegal unless the actual, net effect of the restraint is proven to be anticompetitive. Id. at 42-44. In terms of the synthesis of horizontal restraints jurisprudence just discussed, Respondents appear to argue that this case falls toward the fuller end of the rule of reason spectrum – if not in fact requiring the fullest, or “plenary,” review. To decide whether Respondents are correct, we first must determine whether the agreement between PolyGram and Warner to forgo discounting and advertising of 3T1 and 3T2 falls within the category of restraints that are likely, absent countervailing procompetitive justifi- cations, to have anticompetitive effects – i.e., to lead to higher prices or reduced 48 The Supreme Court has indicated that both sources of insight – the results of case-by-case adjudication and commentary – are relevant as antitrust tribunals form judgments about the competitive significance of observed behavior. State Oil, 522 U.S. at 15. 49 As the Supreme Court said in Socony, “the machinery employed by a combination for price fixing is immaterial.” 310 U.S. at 223. 36 output. In making this assessment, we consider what judicial experience and economic learning tell us about the likely competitive effects of such restrictions.48 A. The Likely Anticompetitive Effects of the Moratorium In keeping with the analytical structure detailed above, we start with an inquiry into whether the restraints at issue here – the agreement not to discount and the agreement not to advertise – are inherently suspect under the antitrust laws, in that they fall within a category of restraints that warrant summary condemnation because of their likely harm to competition. We find ample basis for concluding that they are. 1. The Agreement Not To Discount The anticompetitive nature of the agreement not to discount is obvious. As the ALJ correctly observed, this is simply a form of price fixing, and is presump- tively anticompetitive. See Catalano, 446 U.S. at 648 (agreement to terminate the availability of free credit in connection with purchase of good is “tantamount to an agreement to eliminate discounts, and thus falls squarely within the traditional per se rule against price fixing”); NCAA, 468 U.S. at 100 (horizontal price fixing is “perhaps the paradigm of an unreasonable restraint of trade”).49 Antitrust law’s hostility to price fixing is rooted in uncontroversial economic analysis. As Complaint Counsel’s economic expert, Dr. Stockum, testified, an agreement between competitors not to discount is likely to result in higher prices to consumers, restriction of output, and reduced allocative efficiency. Tr. 583-85; JX 104-B. Dr. Stockum therefore concluded that, absent an efficiency justification, the agreement between PolyGram and Warner not to discount their catalog Three Tenors products was very likely to have had anticompetitive effects. 50 Respondents’ expert witnesses did not testify at trial, and thus were not subject to cross-examination. Our references to the statements of Respon- dents’ experts are to their expert reports and deposition testimony. 51 As the Supreme Court stated in Socony, “the amount of interstate or foreign trade involved is not material . . ., since § 1 of the [Sherman] Act brands as illegal the character of the restraint not the amount of commerce affected.” 310 U.S. at 224 n. 59. 37 Tr. 583-85. Respondents’ own economic expert, Dr. Ordover, agreed that a naked agreement between competitors to restrict price competition has “clearly pernicious effects on competition and consumers.” RX 716 at ¶ 61.50 Moreover, it does not matter that, as Respondents argue, the moratorium applied “only” to two products and “only” for a period of ten weeks.51 It is patently an elimination of a basic form of rivalry between competitors, and properly triggers an obligation by Respondents to come forward with some showing of countervailing procompetitive justification. 2. The Agreement Not To Advertise We also find that the agreement between PolyGram and Warner not to advertise their earlier Three Tenors products is presumptively anticompetitive. The Supreme Court in CDA indicated that, in ordinary commercial markets – like the one at issue here – complete bans on truthful advertising normally are likely to cause competitive harm. 526 U.S. at 773. Indeed, the Court repeatedly has recognized that advertising facilitates competition. By informing consumers of the nature and prices of the goods or services available in a market, and thus creating an incentive for suppliers of the products and services to compete along these dimensions, advertising “performs an indispensable role in the allocation of resources in a free enterprise system.” Bates v. State Bar of Arizona, 433 U.S. 350, 364 (1977); see also Morales v. Trans World Airlines, Inc., 504 U.S. 374, 388 (1992). Restrictions on truthful and nondeceptive advertising harm competi- tion, because they make it more difficult for consumers to discover information about the price and quality of goods or services, thereby reducing competitors’ incentives to compete with each other with respect to such features. See CDA, 526 U.S. at 773 (“restrictions on the ability to advertise prices normally make it more 52 The studies relied on by Dr. Stockum, as well as other empirical literature concerning the impact of advertising restrictions, are in the record at Appendix A to Complaint Counsel’s Findings of Fact, Conclusions of Law, Memorandum of Law in Support Thereof and Order. See Lee Benham, The Effect of Advertising on the Price of Eyeglasses, 15 J.L. & Econ. 337 (1972) (restricting the advertising of eyeglasses raised the average retail price by $7.48); Lee Benham & Alexandra Benham, Regulating Through the Professions: A Perspective on Information Control, 18 J.L. & Econ. 421 (1975) (prices were 25-40% higher in markets with greater professional information controls, including advertising restrictions); Ronald S. Bond et al., Staff Report on Effects of Restrictions on Advertising and Commercial Practice in the Professions: The Case of Optometry (Executive Summary), Bureau of Economics, Federal Trade Commission (Sept. 1980) (price for combined eye exam and glasses was $29 less in cities with least restrictive advertising regimes); John F. Cady, An Estimate of the Price Effects of Restrictions on Drug Price Advertising, 14 Econ. Inquiry 493 (1976) (states restricting the advertising of prescription drugs have prices that are 2.9% higher than states that do not restrict advertising); Steven R. Cox et al., Consumer Information and the Pricing of Legal Services, 30 J. Indus. Econ. 305 (1982) (attorneys who advertised had lower fees than those who did not advertise); Roger 38 difficult for consumers to find a lower price and for [suppliers] to compete on the basis of price”); see also Morales, 504 U.S. at 388; Bates, 433 U.S. at 377-78. These principles apply not just to price advertising, but also to information about qualitative aspects of goods and services. “[A]ll elements of a bargain – quality, service, safety, and durability – and not just the immediate cost, are favorably affected by the free opportunity to select among alternative offers.” Professional Engineers, 435 U.S. at 695. Complaint Counsel’s economic expert testified that an agreement among competitors not to advertise is likely to harm consumers and competition by raising consumers’ search costs and reducing sellers’ incentives to lower prices. Tr. 587-92; JX 104-C. One reason a restriction on advertising may reduce a seller’s incentives to lower prices is that, absent an ability to advertise, lower per- unit prices may not be sufficiently offset by higher volume. Tr. 589-90; JX 105-I ¶ 41; JX 90 at 49-50. Dr. Stockum relied on several empirical studies that have found that advertising restrictions result in consumers’ paying higher prices. Tr. 592-600; JX 104-D (citing studies).52 One of these studies, for example, showed Feldman & James W. Begun, The Welfare Cost of Quality Changes Due to Professional Regulation, 34 J. Indus. Econ. 17 (1985) (total loss of consumer welfare from state regulations governing optometrists that, inter alia, banned price advertising was $156 million); Roger Feldman & James W. Begun, Does Advertising of Prices Reduce the Mean and Variance of Prices?, 18 Econ. Inquiry 487 (1980) (ban on advertising by optometrists and opticians increased prices by 11%); Roger Feldman & James W. Begun, The Effects of Advertising: Lessons from Optometry, 13 J. Hum. Resources 247 (1978) (price is 16% higher in states that ban optometric and optician price advertising); Amihai Glazer, Advertising, Information and Prices – A Case Study, 19 Econ. Inquiry 661 (1981) (grocery prices rose because of newspaper strike in Queens County, NY, that eliminated large amounts of supermarket advertising, and fell after the strike ended); Deborah Haas-Wilson, The Effect of Commercial Practice Restrictions: The Case of Optometry, 29 J.L. & Econ. 165 (1986) (prices were 26-33% lower in markets in which price and non-price media advertising by optometrists occurred); William W. Jacobs et al., Staff Report on Improving Consumer Access to Legal Services: The Case for Removing Restrictions on Truthful Advertising (Executive Summary), Bureau of Economics, Federal Trade Commission (Nov. 1984) (restrictions on attorney advertising resulted in prices that were 5-10% higher); John E. Kwoka, Jr., Advertising and the Price and Quality of Optometric Services, 74 Am. Econ. Rev. 211 (Mar. 1984) (prices of eye exams were $11-$12 lower in markets with advertising than in markets with advertising restrictions); James H. Love & Frank H. Stephen, Advertising, Price and Quality in Self-Regulating Professions: A Survey, 3 Int’l. J. Econ. Bus. 227 (1996) (reviewed 17 studies and found that restrictions on advertising generally have the effect of raising prices paid by consumers); Alex R. Maurizi et al., Competing for Professional Control: Professional Mix in the Eyeglasses Industry, 24 J.L. & Econ. 351 (1981) (advertisers charged approximately $7 less than non-advertisers); Robert H. Porter, The Impact of Government Policy on the U.S. Cigarette Industry, in Empirical Approaches to Consumer Protection Economics 446 (Pauline M. Ippolito & David T. Scheffman eds., 1986) (demand fell by 7.5% as result of 1971 ban on television and radio advertising in the cigarette industry; during the ban, prices increased from 3-6%); John R. Schroeter et al., Advertising and Competition in Routine Legal Service Markets: An Empirical Investigation, 36 J. Indus. Econ. 49 (1987) (advertising made demand more elastic, meaning that consumers were more responsive to price differences); Robert L. Steiner, Does 39 Advertising Lower Consumer Prices?, 37 J. Marketing 19 (Oct. 1973) (advertising resulted in lower toy prices to the consumer). 53 In contrast to the situation in CDA, Respondents here make no argu- ment that the particular industry context renders normal economic conclusions about the competitive impact of price and advertising restrictions inapplicable. This failure is unsurprising, because the present case arises in a conventional commercial context, rather than the professional context that so influenced the Supreme Court’s approach to CDA. See note 32 and accompanying text, supra. In any event, as discussed in Part III.C, infra, the present record amply shows the likely anticompetitive effects of such restraints in the particular context of the recording industry. 40 that even a short-lived restraint on advertising can lead to higher prices. Tr. 599- 600; IDF 247. On the basis of economic theory and empirical studies, Dr. Stockum concluded that, absent an efficiency justification, Respondents’ agreement not to advertise or promote the catalog Three Tenors albums is very likely to be anticompetitive. Tr. 587-92, 616-17; JX 104-D. Dr. Ordover, Respondents’ economic expert, agreed in his deposition that a naked agreement among competitors not to advertise is likely to cause consumer harm. JX 90 at 46- 47. This testimony reinforces the general proposition that restrictions on advertising, such as those imposed here, are likely to reduce competition and harm consumers. B. Respondents’ Justifications Having concluded that both elements of Respondents’ moratorium agreement were indeed inherently suspect restraints of trade because of their likely harm to competition, we turn to Respondents’ proffered justifications. Respondents’ sole argument in this regard is that the moratorium served a plausible procompetitive interest by preventing the PolyGram and Warner operating companies from using the promotional opportunity created by the 1998 Paris concert and the release of the new album to “free ride” on the joint venture.53 In particular, Respondents assert that PolyGram and Warner were concerned that aggressive promotion of 3T1 or 3T2 during the 3T3 release period would divert sales from 3T3, and that the prospect of such diversion could induce them to withhold promotional efforts in support of 3T3. They further assert that lack of 54 Respondents also assert, in passing, that the moratorium prevented the PolyGram and Warner operating companies from using “confidential marketing plans developed by the joint venture partners.” Respondents’ Opening Brief at 44. However, Respondents do not develop this argument further and cite to no record evidence indicating that the moratorium was intended to protect against the misuse of confidential marketing plans. 55 Although Respondents state their justification for the moratorium in various ways, their arguments all amount to the same thing: that restraining competition from 3T1 and 3T2 enhanced the marketing of the new joint venture product. 41 success with 3T3 could have undermined the success of subsequent joint venture products – i.e., a proposed “Greatest Hits” album and a Boxed Set.54 We reject these arguments as a matter of law because they go far beyond the range of justifications that are cognizable under the antitrust laws.55 Respondents are not asserting that restraints on the joint venture activities are reasonably necessary to achieve efficiencies in its operations, nor even that expansion of the joint venture is reasonably necessary to achieve such efficiencies. Rather, they are arguing that competitors may agree to restrict competition by products wholly outside a joint venture, to increase profits for the products of the joint venture itself. Such a claim is “nothing less than a frontal assault on the basic policy of the Sherman Act,” Professional Engineers, 435 U.S. at 695, for it displaces market- based outcomes regarding the mix of products to be offered with collusive deter- minations that certain new products will be offered under a shield from direct competition. Preventing free-riding can be a legitimate efficiency. The most widely recognized application in antitrust of this efficiency is, as Respondents suggest, limiting intrabrand competition to improve interbrand competition. See Conti- nental T.V., Inc. v. GTE Sylvania Inc., 433 U.S. 36, 54-55 (1977). In such cases, the scope of the restraint is necessarily limited to products that are within the control (at least initially) of the entity that owns the restricted brand. Here, despite Respondents’ invocation of a Three Tenors “brand,” there is obviously no such thing, because one entity did not legally control all Three Tenors products. The 56 Had this case involved a merger to create a single entity with rights to market all Three Tenors products, a different analysis would have been required – i.e., one that would weigh potential anticompetitive effects against the prospect of integrative and other efficiencies, under the standards of Section 7 of the Clayton Act, 15 U.S.C. § 18. 42 marketing rights to 3T1 and 3T2 were held not by the joint venture but, rather, independently by the parties to the venture. RX 716 ¶ 31. See supra Part I.C.56 Respondents draw our attention to cases in which courts have declined to condemn restrictions that co-venturers have imposed upon each other when the restrictions were justified, at least in part, as reasonable means to control free- riding by the co-venturers. These cases are readily distinguishable from the case at hand. The restraints upheld in the “free-riding” cases Respondents rely upon were limited to the products of the joint venture or other single economic entity involved. For example, in Polk Bros., Inc. v. Forest City Enterprises, Inc., 776 F.2d 185 (7th Cir. 1985), two retail chains whose offerings were largely comple- mentary, but which were at least potential competitors, agreed to open a new store offering, side by side, the full range of their goods. To protect their respective economic interests and make the new venture possible, they agreed to refrain from carrying competing goods at that location. 776 F.2d at 187. The venturers did not agree to restrict competition between their other stores. Id. In Rothery Storage & Van Co. v. Atlas Van Lines, Inc., 792 F.2d 210 (D.C. Cir. 1986) (“Rothery”), cert. denied, 479 U.S. 1033 (1987), Atlas, a national van line that contracted with numerous local agent-carriers, altered its previously more flexible arrangement by generally requiring that any moving company doing business as its agent cease interstate carriage on its own account and provide such carriage exclusively in conjunction with Atlas (although competition by wholly independent affiliates was allowed in some circumstances). 792 F.2d at 213, 217. Atlas’s restriction simply required agent-carriers to bring within the integrated joint venture all of their interstate carriage that used Atlas’s equipment, uniforms, services, or other assets of the Atlas network. Because Atlas demonstrated that this restraint was reasonably necessary to eliminate free-riding and thus preserve the efficiencies of the joint venture and because Atlas had only a small percentage of the overall national market, the court upheld the restraints under the rule of reason. Id. at 229. 57 Prior to the moratorium agreement, these independent entities had planned to conduct marketing campaigns for 3T1 and 3T2 during the release of 3T3. IDF 102-05, 115-18. Moreover, Respondents were concerned that it would be difficult for PolyGram and Warner to implement the moratorium consistently on a worldwide basis, because they did not have complete control over the prices for 3T1 and 3T2 charged by their operating companies. IDF 126. Ultimately, however, PolyGram and Warner succeeded in enforcing the moratorium. See supra Part I.C. 58 As discussed in Part III.C.3., infra, the record reveals that this phenomenon is common in the music industry. JX 91 at 126-27; JX 97 at 46; CX 609 at 71-73, 83-84; CX 610 at 52-54. It is common in many other industries, as well. 43 In the present case, however, Respondents and Warner did not bring all of their Three Tenors products into a single, integrated joint venture; indeed, the joint venture agreement expressly provided that PolyGram and Warner could continue to exploit 3T1 and 3T2. JX 10-V. Nor did Respondents and Warner limit the restrictive effects of the moratorium to the product within the joint venture – i.e., 3T3. Rather, they left each of the three Three Tenors products in the hands of an independent economic entity, yet agreed to restrict competition by two of those entities – Respondents with respect to the marketing of 3T1 and Warner with respect to the marketing of 3T2.57 Thus, the issue here is whether a joint venture can claim the “efficiency” of limiting “free-riding” from competing products the joint venture neither owns nor otherwise legally controls. The sort of behavior that Respondents disparage as “free-riding” – i.e., taking advantage of the interest in competing products that promotional efforts for one product may induce – is an essential part of the process of competition that occurs daily throughout our economy. For example, when General Motors (“GM”) creates a new sport utility vehicle (“SUV”) and promotes it, through price discounts, advertising, or both, other SUVs can “free ride” on the fact that GM’s promotion inevitably stimulates consumer interest, not just in GM’s SUV, but in the SUV category itself.58 Our antitrust laws exist to protect this response, because it is in reality the competition that drives a market economy to benefit consumers. There is no doubt that GM’s SUV will likely be more profitable if its competitors do not respond. Promoting profitability, however, is not now, nor has 59 The Catalano Court stated: [I]n any case in which competitors are able to increase the price level or to curtail production by agreement, it could be argued that the agreement has the effect of making the market more attractive to potential new entrants. If that potential justifies horizontal agreements among competitors . . . it would seem to follow that the more successful an agreement is in raising the price level, the safer it is from antitrust attack. Nothing could be more inconsistent with our cases. 446 U.S. at 649. 60 As mentioned above, see supra Part I.C., Sony released a Three Tenors Christmas album in 1999. 61 The transcript of the oral argument reads “per se legal” (Transcript of Nov. 4, 2002 Oral Argument at 74:24), but it is clear from the surrounding discussion of the Sony hypothetical that Respondents’ counsel actually said (or meant) “per se illegal.” 44 it ever been, recognized as a basis to restrain interbrand competition under the antitrust laws. See Catalano, 446 U.S. at 649;59 Law, 134 F.3d at 1023 (“mere profitability or cost savings have not qualified as a defense under the antitrust laws”); Chicago Prof’l Sports Ltd. Partnership v. National Basketball Ass’n, 754 F. Supp. 1336, 1359 (N.D. Ill. 1991), aff’d, 961 F.2d 667 (7th Cir. 1992). During the oral argument, Respondents in effect conceded this flaw in their argument in their response to a hypothetical positing that Sony had received the rights for 3T3 and then Sony had entered into the same moratorium agreement with Warner and PolyGram restricting price discounting and advertising of 3T1 and 3T2 during the 3T3 release period.60 This hypothetical assumes that the same benefits to the Three Tenor “brand” exist that Respondents assert exist in their joint venture. Respondents conceded that for Sony to enter into such an agreement with Warner and PolyGram would be per se illegal,61 even if it might maximize the value of the Three Tenors “brand” in the long term. Transcript of Nov. 4, 2002 Oral Argument at 74-75. Although Respondents claim that the Sony hypothetical 62 The Commission’s decision in 1984 to permit General Motors and Toyota to engage in a production joint venture provides an instructive point of comparison. General Motors Corp., 103 F.T.C. 374 (1984). No feature of the GM-Toyota joint venture, either as proposed by the parties or as ultimately approved by the Commission, restricted competition between the two firms concerning existing automobile models that they previously had developed independently. This is a critical distinction between that case and the present one. As a leading commentator noted, “[w]hat excuses the GM-Toyota venture from charges of per se unlawful price fixing is that the venturers did not enter into any agreement to fix the price of their nonventure output.” XI Hovenkamp, Antitrust Law ¶ 1908e, at 237-38. 45 is inapposite because the parties here were engaged in a joint venture and own the competitive products, they provided no principled reason why this distinction should make a difference. In each, three products are offered, by three different and independent economic entities. In each, the competitive efforts on behalf of two products are restricted to shield a third product from competition. In each, there is a blatant departure from the principles of free competition on which our antitrust laws are based. Nor does the fact that the shielded product is a new introduction to the market justify such market manipulation. Suppose, to return to our SUV example, that GM and one of its rivals enter into a joint venture to produce a new SUV, and the parties restrict the competition from their existing, non-joint-venture SUVs to protect the market for the new SUV.62 Any argument that such a stifling of competition is “necessary” to bring the new product to market would face the same fundamental problem that condemns Respondents’ arguments here. Although the antitrust laws favor product innovation, the very concept of a free market is that competitive forces themselves will induce the production of new products that consumers desire and whose availability will enhance consumer welfare. “Antitrust law presumes that competitive markets offer sufficient incentives and resources for innovation, and that cartel pricing leads not to a dedication of newfound wealth to the public good but to complacency and stagnation.” Freeman v. San Diego Ass’n of Realtors, 322 F.3d 1133, 1152 (9th Cir. 2003). If a “new” product can succeed in a free marketplace only if it is shielded from competitive forces by a facially anticompetitive agreement between existing competitors, then it is likely no loss to consumers if it is not introduced. 63 Respondents’ reliance on Example 10 in Section 3.36(b) of the Collaboration Guidelines is misguided. That example addresses the analysis of restrictions imposed by co-venturers in the development of new word processing software products – including, potentially, the cessation of sales of preexisting, competing products. The example makes clear, however, that such restraints may be justified only if they achieve “cognizable efficiency goals.” Id. (emphasis added). Specifically, the example indicates that such restrictions might be justi- fied if they were necessary for the activities of the joint venture itself, as for monitoring the venturers’ contributions of assets or preventing one participant from misappropriating assets the other contributed. The example does not support the notion that a restraint on the marketing of non-venture products can be justi- fied simply because it would increase sales opportunities for the joint venture product. On the contrary, the Guidelines make clear that claims “premised on the notion that competition itself is unreasonable . . . are insufficient as a matter of law.” Collaboration Guidelines, at § 3.2. Moreover, as discussed in Example 9 of the Guidelines, cost savings from depriving consumers of information useful to their decision making (like the advertising restrictions at issue here) amounts to a service reduction, not a cognizable efficiency. Further, unlike the joint venture in Example 10, the collaboration at issue here was merely a marketing venture. PolyGram and Warner did not create a novel product. They did not produce the 1998 Three Tenors concert; that was done independently by concert promoter Rudas. Instead, PolyGram and Warner merely collaborated to distribute the audio and video recordings of the 1998 concert. See supra Part I.C. 46 To allow such an “efficiency” to justify an agreement between competitors to restrict promotion of competing products is to displace market forces with collusive decisions by competitors regarding what new products consumers ought to be offered.63 Indeed, the argument Respondents advance here is remarkably similar to a justification that the NCAA Court considered and rejected as antithetical to the antitrust laws. There, addressing the NCAA’s argument that restrictions on television broadcasts of college football games were necessary to protect live attendance at games, the Court stated: 47 At bottom the NCAA’s position is that ticket sales for most college games are unable to compete in a free market. The television plan protects ticket sales by limiting output – just as any monopolist increases revenues by reducing output. By seeking to insulate live ticket sales from the full spectrum of competition because of its assumption that the product itself is insuf- ficiently attractive to consumers, petitioner forwards a justification that is inconsistent with the basic policy of the Sherman Act. NCAA, 468 U.S. at 116-17. See also Professional Engineers, 435 U.S. at 696 (“[T]he Rule of Reason does not support a defense based on the assumption that competition itself is unreasonable. Such a view of the Rule would create the ‘sea of doubt’ on which Judge Taft refused to embark in Addyston, 85 F. at 284, and which this Court has firmly avoided ever since.”). Another way of analyzing this issue is that the restraints here are not “ancillary” to the production of efficiencies in the sense that Sherman Act cases have employed that concept, even assuming (contrary to our conclusion in Part III.C.3, infra) that, as a factual matter, restricting the marketing of 3T1 and 3T2 was reasonably necessary to ensure the vigorous marketing of 3T3. To qualify as an “ancillary” restraint, “an agreement eliminating competition must be subordi- nate and collateral to a separate, legitimate transaction,” and it must also “be related to the efficiency sought to be achieved.” Rothery, 792 F.2d at 224. A determination of ancillarity includes, of course, the factual inquiry whether a particular restraint was indeed reasonably necessary to permit the parties to achieve a particular efficiency. See infra Part III.C.3. But that factual inquiry is not the only pertinent consideration. Suppose, for example, General Motors and Toyota asserted that, to provide incentives for marketing of a new solar-powered car, they would eliminate price promotions on their conventional vehicles. Such an argument would be rejected because it is not sufficiently “related to” the efficiency to be furthered. Cases in which defendants successfully invoked the doctrine of ancillary restraints consistently have involved restraints that affect the joint venture at issue, but not products outside its scope. This was true in both Rothery and Polk 64 As discussed in Part III.C.3, infra, the restraints on the marketing of 3T1 and 3T2 also fail to qualify as ancillary as a matter of fact, in that the record shows that such restrictions were not actually necessary to ensure the introduction and vigorous promotion of 3T3 and any covered follow-on products. 65 Accordingly, we have no need to determine whether Respondents’ proffered justification is “plausible” in a purely factual sense. Because it is not 48 Brothers, as discussed above. Similarly, in BMI, the Court upheld the joint setting of prices for the joint venture product (blanket music licenses) because it “accompanie[d] the integration of sales, monitoring, and enforcement against unauthorized copyright use.” 441 U.S. at 20. Significantly, the pricing arrangement approved in BMI did not include products outside the joint venture – i.e., licenses on individual compositions – which remained available and were not subject to restraints. Id. at 23-24; see XI Hovenkamp, Antitrust Law ¶ 1908e, at 237-38. Respondents have not cited any cases, nor are we aware of any, in which restraints on the sales of non-joint-venture products have been upheld as “ancillary” to the production of efficiencies by the joint venture itself. On the contrary, the Commission has long recognized that restraints on activities “outside the ambit of the joint venture” cannot be hidden under its cloak. See Brunswick Corp., 94 F.T.C. 1174, 1277 (1979), aff’d sub nom. Yamaha Motor Co., Ltd. v. Federal Trade Commission, 657 F.2d 971, 981 (8th Cir. 1981), cert. denied, 456 U.S. 915 (1982). In the present case, Respondents and Warner chose to retain control over their respective existing Three Tenors products and to form a joint venture limited to 3T3 and specified follow-on products (i.e., a possible “Greatest Hits” recording and a Boxed Set). They cannot claim the integrative efficiencies that could conceivably have been brought about by combining the production and marketing of all Three Tenors products. Accordingly, the restrictions on the marketing of 3T1 and 3T2 cannot be considered “ancillary” to the present joint venture, as a matter of law, because they are not related to the efficiencies the joint venture was created to produce.64 Thus, we hold that the Respondents’ “free-riding” argument is simply an attempt to shield themselves from legitimate interbrand competition. As such, the proffered justification is not cognizable under antitrust law.65 This conclusion, cognizable under antitrust law, it has no relevance to our analysis. See note 42, supra. In any event, as we determine in Part III.C.3, infra, Respondents’ attempted defense also fails factually. 66 See discussion of NCAA, at p. 19-21, supra; see also Professional Engineers, 435 US at 688 ("to evaluate this argument it is necessary to identify the contours of the Rule of Reason and to discuss its application to the kind of justification asserted by petitioner") and at 435 U.S. at 695 ("It is this restraint that must be justified under the Rule of Reason . . ."). Of course, even this type of analysis is unnecessary in cases with no possible arguments that restraints are needed to achieve beneficial results, and a more traditional per se approach remains appropriate. See, e.g., United States v. Andreas, 39 F. Supp. 2d 1048, 1058-61 (N.D. Ill. 1998) (rejecting arguments that rule of reason can apply to criminal case charging price fixing and volume allocation imposed to restrict output), aff’d, 216 F.3d 645, 666-68 (7th Cir. 2000). Such matters are commonly the subject of criminal prosecution and are appropriately deemed per se illegal, as are other restraints for which the proffered justifications can likewise be dismissed summarily. See also Palmer, 498 U.S. at 49-50; SCTLA, 493 U.S. at 424; Catalano, 446 U.S. at 649-50. 49 together with our previous conclusion that the restraints at issue are of the sort that are likely to harm competition, provides us with ample ground to condemn Respondents’ actions as unlawful under Section 1, without further analysis. Arguably, this conclusion could be characterized as a finding of “per se illegality” in that we conclude that the restraints at issue are “naked” restraints on competi- tion because they lack a cognizable justification. Yet our mode of analysis, in which we evaluate the proffered justifications at some length and ultimately reject them as not cognizable in an antitrust analysis, closely tracks that of the Supreme Court in Professional Engineers and NCAA – both cases that the Court described as applying the rule of reason.66 In the end, the label matters less than the substance of the analysis, the purpose of which remains “to form a judgment about the competitive significance of the restraint.” Professional Engineers, 435 U.S. at 692. Here, we have no doubt that the restraints before us harm competition and must be condemned. 67 Cooperative advertising is a monetary commitment that the record label makes to retailers to support both out-of-store advertising (e.g., print, radio, and television advertising) and in-store promotion (e.g., posters and floor dis- plays). Out-of-store advertising is intended to draw customers into the store by informing them where a recording may be purchased and at what price. In-store advertising is designed to increase the likelihood that, once inside the store, the consumer buys a specific recording. JX 105-F; Tr. 48-54, 58-60, 194-96. When PolyGram provides cooperative advertising funds, the retailer provides the advertising and then deducts the value of the cooperative advertising from the amount it pays for the product it purchases from PolyGram. Cooperative advertis- ing thus functions as a price discount. IDF 217-18. Indeed, industry participants 50 C. A More Detailed Factual Analysis Our analysis could properly end at this point. Respondents’ only proffered justification is not cognizable as a matter of law, and therefore triggers no need to go beyond the analysis presented above. Even if we concluded, however, that Respondents had offered a cognizable and plausible justification and that a more elaborate analysis were therefore needed, analysis of the facts here would only reconfirm our ultimate conclusion. The extensive factual record regarding practices in the recording industry and Respondents’ own prior course of conduct establishes that the harm to competition not only is inferable from the nature of the conduct but is established as a matter of fact. And the record likewise shows that Respondents’ proffered justification regarding free riding and the supposed need to ensure the vigorous promotion of 3T3 would fail as a factual matter, even if it were legally cognizable. 1. Competitive Effect of Respondents’ Discounting Restrictions The record evidence shows that the moratorium’s price restraint not only was inherently suspect, but also actually harmed competition and consumers. In the sale of recorded music, as in other industries, price discounting is an important dimension of competition. IDF 238-42. Executives from PolyGram and Warner testified that their companies commonly offer price discounts to retailers, on catalog products as well as new releases, and that such discounts increase sales. IDF 239. PolyGram and Warner also commonly provide retailers with cooperative advertising funds, which function as a discount from the wholesale price.67 IDF recognize that cooperative advertising funds are a form of discount, because they represent the partial assumption by the recording company of expenses that retailers would otherwise bear. See CX 603-P (in camera) (discussion by Warner of cooperative advertising). 51 217-18; CX 603-Z-18 (in camera). These wholesale discounts encourage retailers to sell the product to consumers at reduced retail prices. IDF 220; JX 100 at 91-92 (in camera). Prior to the moratorium, Respondents discounted prices as part of the marketing strategy for their respective Three Tenors products. In 1994, PolyGram responded to the release of 3T2 by launching an aggressive marketing campaign for 3T1 worldwide, with price discounting in many markets. JX 29 (“PolyGram were able to sell an additional one million copies of their 1990 album on the back of our new record in 1994. This was achieved through aggressive TV advertising, print advertising, extensive rack exposure of their record at retail and a price reduction.”) (emphasis added); JX 12 (in the U.S., 3T1 audio sales in 1994 increased 274% over 1993 sales as a result of marketing campaign); IDF 214-21. In the United States, for example, PolyGram provided cooperative advertising funds to retailers to increase sales and encourage lower retail prices for 3T1. IDF 219-20. In 1996 and 1997, during the Three Tenors’ world concert tours, PolyGram again offered 3T1 at a discounted price in many markets. IDF 224-25, 241; CX 299 at 3TEN00005903 (“You can be certain Decca will be planning to exploit this concert tour with pricing campaigns . . . .”). In early 1998, many PolyGram and Warner operating companies planned to reduce the price of 3T1 and 3T2 as part of aggressive marketing campaigns, including promotional activities planned for the weeks surrounding the release of 3T3. IDF 102-05, 115- 18. As a result of the moratorium agreement, however, 3T1 and 3T2 ultimately were sold only at full price during the release of 3T3. IDF 170-81. Respondents argue there is no evidence that the pricing (or advertising) of 3T1 or 3T2 in the United States would have been different without the moratorium. In particular, Respondents assert that in 1994 PolyGram did not discount 3T1 in the United States, and that evidence cited by the ALJ regarding PolyGram’s and Warner’s plans in 1998 to discount 3T1 and 3T2 related solely to operating companies outside of the United States. Respondents’ Opening Brief at 16-17. Respondents appear, however, to hold an artificially narrow view of what 68 The evidence is clear that PolyGram employed cooperative adver- tising for 3T1 in 1994 in the United States. For example, in September 1994 – the first full month after the release of 3T2 – PolyGram returned to retailers through 3T1 cooperative advertising programs approximately 9% of the money generated from 3T1 sales. IDF 219. 52 constitutes price discounting. Although one method of price discounting, called a “mid-price campaign,” is not used in the United States, Tr. 184-86, the evidence shows that record companies in the United States – including PolyGram and Warner – routinely use other forms of price discounting, such as wholesale discounts offered to retailers on new releases or restocking campaigns for catalog products. JX 100 at 91-92 (in camera); CX 609 at 49-50; Tr. 44-45. Record companies in the United States – again, including PolyGram and Warner – also use cooperative advertising to achieve what is effectively a discount in the wholesale price, without actually lowering the suggested list price. Tr. 66-68, 187, 808; IDF 217-18, 220.68 Moreover, the moratorium applied worldwide, not merely to foreign markets. As Dr. Stockum explained, when direct competitors form an agreement not to discount, “it is a safe economic inference to draw that they intend to stop discounting that would otherwise have occurred.” JX 85 at 45-46. This inference is particularly safe where it appears that the parties’ counsel cautioned them about the legal risks of a moratorium on discounts. See p. 9, supra. On this record, we find that the agreement by PolyGram and Warner not to discount 3T1 and 3T2 in the period surrounding the release of 3T3 not only is presumptively anticompetitive, but also eliminated actual price discounting that had occurred previously in the industry, including competition between 3T1 and 3T2 upon the release of 3T2. 2. Competitive Effect of Respondents’ Advertising Restrictions Here, in contrast to CDA, Respondents made no effort to articulate any reason why the market in question (the sale of recorded music) falls outside the “general rule” that advertising restrictions tend to have anticompetitive effects. See 526 U.S. at 771. Nevertheless, the record evidence confirms that such principles indeed apply fully to the recorded music industry, and that the advertising restrictions imposed here were harmful to competition. See Tr. 601- 03. The record shows that advertising is an important basis of competition in this 69 For example, Warner’s 1994 marketing plan for 3T2 stated: In order to counter the perceived threat of competitive imitation products which will aim to satisfy demand in the period directly around the concert using similar repertoire and perceptually identical artists, the concept of the genuine or “real thing” will underpin all local implementation of the [marketing strategy]. CX 259 at 3TEN00011109. 53 industry. JX 105-F-G. Record companies spend considerable sums of money advertising their products. CX 609 at 57-59; JX 101 at 12-13. Such advertising serves to inform consumers about the availability of alternatives, sales locations, prices, and quality differences among competing products. Tr. 53-54, 58-59, 62- 64. Complaint Counsel’s music industry marketing expert, Dr. Moore, explained that a record company’s decisions regarding advertising and wholesale price are linked, and if there is no advertising, there is less incentive for the company to offer the recording at a significantly reduced price. JX 105-I ¶ 41. Dr. Moore further testified – and Respondents’ executives confirmed – that record companies advertise to increase their sales, and that such advertising generally results in lower retail prices for consumers. Tr. 58-59, 64-67; JX 87 at 79-80, 90; CX 609 at 59; CX 610 at 50. Furthermore, before the moratorium, advertising was an important part of competition between 3T1 and 3T2. In 1994, when 3T2 was released, PolyGram advertised to inform consumers that 3T1 was the “original” Three Tenors record- ing, was still widely available, and indeed was often available at a discounted price. IDF 210-20. Largely as a result of its marketing campaign, PolyGram sold almost one million audio and video recordings of 3T1 in the second half of 1994, as compared with 377,000 in the same period in 1993. JX 12. In turn, Warner used advertising to create a distinct identity for 3T2, suggesting to consumers that the newer release was the superior product. IDF 201-09.69 PolyGram and Warner again used advertising to highlight the advantages of their respective Three Tenors products during the Three Tenors’ world concert tours in 1996 and 1997. IDF 224-34. 54 In 1998, PolyGram and Warner operating companies began to plan advertising campaigns for their respective catalog Three Tenors products in connection with the upcoming Paris concert. IDF 102-03, 105, 115-18, 255-58. PolyGram and Warner subsequently instructed their operating companies that, because of the moratorium agreement, advertising of 3T1 and 3T2 had to end before 3T3 was released. IDF 107, 147-49. The ban on advertising was intended to protect sales of 3T3 by withholding information from consumers about the nature and price of competing products. As one Warner executive explained at trial, the companies did not want consumers to “start comparing the repertoire along with the price and make a determination that, you know, the ‘94 concert is just fine for a few dollars less.” Tr. 487. We agree with the ALJ that the anticompetitive effect of this strategy is obvious. IDF 224-34. 3. Inadequacy of Respondents’ Free-Rider Defense The foregoing analysis shows that the price and advertising restrictions Respondents imposed were inherently suspect as a matter of economic theory and also were demonstrably anticompetitive in the particular industry context in which they were imposed. Although we have found it unnecessary to engage in “the fullest market analysis,” CDA, 526 U.S. at 779, we have examined evidence of industry practice and the past practices of the very participants in the present scheme, as well as the consistent economic literature regarding the likely effects of such practices. By any standard, this is an enquiry “meet for the case,” allowing us to arrive at a “confident conclusion” about the anticompetitive nature of these restraints. Id. at 781. An antitrust defendant can avoid liability in these circumstances only by making a concrete showing of “countervailing procompetitive virtue.” See IFD, 476 U.S. at 459. Respondents have failed to make such a showing. As discussed above, Respondents’ only proffered justification is imper- missible as a matter of law, because the supposed “efficiency” of restraining competition in the offering of products outside of a joint venture to enhance market opportunities for a new joint venture product is not cognizable under the antitrust laws. Nevertheless, in this section we examine the record evidence on these restraints and conclude that, even if Respondents could properly defend on the basis that restricting the marketing of 3T1 and 3T2 was reasonably necessary 55 to ensure the vigorous marketing of 3T3, the record simply does not support that argument as a factual matter. The joint venture unquestionably would have proceeded and the new product would have been brought to market without the moratorium. Initially, Warner planned to market and distribute 3T3 on its own, without any collabora- tion from PolyGram. IDF 52. Furthermore, PolyGram and Warner were con- tractually committed to the formation of the joint venture and the creation of 3T3 months before discussions of the moratorium began. IDF 263. Although the timing of the moratorium is not dispositive, it is certainly relevant to an assess- ment of whether the moratorium was reasonably necessary to achieve the procom- petitive benefits of the collaboration. At trial, a Warner executive testified that even if PolyGram and Warner had not agreed to the moratorium, Warner was committed to distribute 3T3 in the United States. Tr. 446-47. Moreover, the fact that the joint venture agreement itself expressly contemplated that PolyGram and Warner would remain free to exploit the earlier Three Tenors albums strongly suggests that the parties did not view a ban on competition from these products as important to the efficient operation of the joint venture. JX-10-J-K. The evidence in this case shows that the prospect that PolyGram and Warner operating companies would discount and advertise 3T1 and 3T2 during the 3T3 release period did not diminish Warner’s incentives to promote 3T3 in the United States. Respondents’ marketing expert, Dr. Wind, acknowledged in his deposition that firms commonly capitalize on the promotional activities of their competitors, and sellers generally respond to this challenge by using advertising and other marketing tools to create a distinct identity for the target product. JX 91 at 125-29, 133-34; IDF 277-79. In particular, within the recorded music industry, the diversion of sales identified by Respondents is commonplace, and advertising intended to benefit one album often leads to sales of competing albums, including catalog albums by the same artist. IDF 280; Tr. 87-88, 264-65; JX 89 at 33-35; JX 87 at 69-72; JX 101 at 183-84; JX 102 at 114-15; JX 609 at 71- 73. As the president of WMI wrote when informed that the moratorium agreement would prevent his operating companies from implementing their plans to promote and discount 3T2 when 3T3 was released: 70 In economic terms, one reason for this practice is that, for certain consumers, prior recordings are apparently complements, not substitutes. That is, for these consumers a new recording can increase the attractiveness of previous recordings. 56 There is nothing sinister nor underhanded in marketing catalog on the back of a significant related event or new release. In fact, as you well know, this is the normal and traditional practice of our industry. JX 8.70 Complaint Counsel’s music industry marketing expert testified, and the parties’ executives confirmed, that the prospect of a new album’s losing sales to competing catalog products typically does not lead record companies to curtail their marketing of a new album. Tr. 88-90; JX 105-H; CX 610 at 54-55; CX 609 at 71-80, 85-86. For example, when Warner released 3T2 in 1994, it anticipated that PolyGram would take advantage of the promotional opportunity arising from the release of 3T2 to advertise and discount 3T1. IDF 202. But Warner did not cut back on its marketing of 3T2. To the contrary, it launched an aggressive and expensive international marketing campaign in support of 3T2, competing by creating a distinct identity for 3T2. Tr. 89-98; IDF 201, 203-09. The evidence here shows that marketing activities in support of 3T3 would not have been curtailed on account of the promotion of 3T1 and 3T2. IDF 288-91. Witnesses representing both Warner and PolyGram testified that 3T3 would have been appropriately promoted without the moratorium, and that the moratorium had no effect on the resources for advertising and promoting 3T3. Tr. 490; JX 94 at 87-89; JX 95 at 89-90; JX 101 at 85-86; IDF 288-91. Indeed, in June 1998, when it appeared that the moratorium would fall apart, PolyGram did not alter its marketing strategy or cut back on its advertising budget. IDF 129. Respondents fail to point to any convincing countervailing evidence that “opportunistic” behavior by PolyGram and Warner operating companies would have led Warner to reduce its level of marketing of 3T3 in the United States. Even Respondents’ economic expert, Dr. Ordover, was unable to conclude that promo- tion of 3T1 and 3T2 was a significant concern in the United States; rather, he found that the moratorium was motivated by concerns about promotion of 3T1 and 57 3T2 in Europe. JX 90 at 36-37; IDF 294-96. Even if the evidence supported a conclusion that promotional activities by the operating companies in Europe were a concern, this would not justify a ban on discounting and advertising in the United States. See Rothery, 792 F.2d at 224 (“If [a restraint] is so broad that part of the restraint suppresses competition without creating efficiency, the restraint is, to that extent, not ancillary.”). Moreover, although Dr. Ordover opined that the moratorium was “reasonably necessary” to avoid free-riding, he defined “reasonably necessary” as meaning not obviously pretextual. IDF 297-98. This meaning of “reasonably necessary” is contrary to the case law. See Rothery, 792 F.2d at 224 (restraint “must be subordinate and collateral to a separate, legal transaction” and “related to the efficiency sought to be achieved”). Dr. Wind, Respondents’ marketing expert, opined that the moratorium plausibly benefitted consumers because it provided incentives for PolyGram and Warner to produce 3T3 and invest in promoting the album, but he could not identify any record evidence that supported his opinion. JX 91 at 111-15, 117-18. Accordingly, we agree with the ALJ that the opinions of Respondents’ experts are entitled to little weight. ID at 58-59, n. 25. Respondents also fail to point to any convincing evidence to support their contention that the moratorium increased the likelihood that the parties would release a Three Tenors Boxed Set and a Greatest Hits album. Although aggressive promotion of 3T1 and 3T2 during the launch of 3T3 might have diverted some sales of 3T3 to the other products (with consumers benefitting from lower prices), presumably there would have been at least as many total units sold during that period. This scenario may well have been less profitable for the joint venture, but it is not apparent that the parties’ possible decision in the future to release these additional Three Tenors products would have depended on achieving greater sales of 3T3, as opposed to sales of 3T1 or 3T2. A Warner executive testified that the decision whether to release a Greatest Hits album was not related to a moratorium on price discounting, and that, as of early 2001, the disappointing sales of 3T3 had not dissuaded Warner and PolyGram from planning to release a Three Tenors Boxed Set or a Greatest Hits album. JX 101 at 76, 110-11, 113-15. See also JX 24 (“PolyGram has insisted . . . on having these box set and ‘greatest hits’ rights in order to ‘hedge their bets’ and give us an additional source of income in case the 1998 album does not perform up to expectations.”). 71 Respondents’ argument about consumer confusion – that eliminating the “clutter and confusion” of competing products was “in the customer’s best interest,” JX 94 at 80 (Saintilan Dep.) – is similar to a justification that the Supreme Court rejected in IFD. See p. 22, supra. 58 At most, Respondents’ record citations suggest that some PolyGram and Warner executives harbored vague concerns that discounting and advertising of 3T1 and 3T2 during the launch of 3T3 might have “devalued” the Three Tenors “brand” (jeopardizing future demand for Three Tenors products) or resulted in customer confusion (leading customers to purchase a different album than intended or perhaps not purchase anything at all). JX 89 at 57-58; JX 94 at 80-82. Respondents, however, offer no evidence indicating that these are valid concerns.71 In 1994, PolyGram responded to the release of 3T2 by discounting and aggressively promoting 3T1; and during the Three Tenors world tours in 1996 and 1997, both companies mounted promotional campaigns, which included discounting in many markets. See supra Part I.C. There is no evidence that any of these promotional activities “devalued” the Three Tenors “brand,” unduly con- fused consumers, or otherwise threatened Three Tenors output. IV. REMEDY Having found a violation of Section 5 of the FTC Act, the Commission is empowered to enter an appropriate order to prevent a recurrence of the violation. The Commission has wide discretion in its choice of a remedy. Federal Trade Commission v. National Lead Co., 352 U.S. 419, 428 (1957); Jacob Siegel Co. v. Federal Trade Commission, 327 U.S. 608, 611 (1946). “[T]he Commission is not limited to prohibiting the illegal practice in the precise form in which it is found to have existed in the past,” but “must be allowed effectively to close all roads to the prohibited goal, so that its order may not be by-passed with impunity.” Federal Trade Commission v. Ruberoid Co., 343 U.S. 470, 473 (1952). The remedy selected, however, must be reasonably related to the violation found to exist. Id.; Jacob Siegel, 327 U.S. at 613. The order we issue narrowly prohibits the Respondents from engaging in the conduct that we have concluded was unlawful without impeding their ability to engage in legitimate joint venture activity. Paragraph II of the order requires Respondents to cease and desist from entering into an agreement with a competitor 72 Respondents claim, without citing authority for the proposition, that this provision improperly reverses the substantive and procedural burdens under the antitrust laws. We disagree. Requiring Respondents to demonstrate a justification for conduct that is inherently suspect is consistent with the analytical framework set forth in the relevant cases and followed in this opinion. 59 to fix prices of, or restrict truthful or “nondeceptive” advertising for any audio or video product in the United States. Paragraphs III.A. and III.B. specifically provide that the order does not prohibit Respondents from entering into a written agreement to set prices of or restrict the advertising for any audio or video product if the agreement is reasonably related to a lawful joint venture and reasonably necessary to achieve its procompetitive benefits. Paragraphs III.C. and III.D. provide that the order does not prohibit Respondents from entering into a written agreement to set prices of or restrict the advertising for any jointly produced audio or video product. Paragraph III.E. provides that Respondents are not prohibited from complying with an industry code or ethical standard intended to restrict the marketing to children of audio and video products rated with a parental advisory. Paragraph III.F. provides that, in any action by the Commission alleging a viola- tion of this order, the burden is on the Respondents to show that the challenged conduct satisfies the conditions of Paragraphs III. A-E.72 These provisions are clearly related to the law violation found to exist and no broader than necessary to prevent a recurrence of the violation. Paragraphs IV, V, VI, VII, and VIII set forth Respondents’ compliance obligations under the order. We have altered the ALJ’s proposed order by shortening Respondents’ reporting obligations under Paragraph IV.B. from nine to five years. These provisions are designed to assist the Commission in monitoring compliance with the order, and they impose a small burden on Respondents. Respondents argue that a cease and desist order is not supported in this case because there is no threat that similar conduct will recur. We disagree. The marketing challenge that gave rise to the Three Tenors moratorium – i.e., the fear that a new release by one of Respondents’ recording artists may lose sales to the artist’s older albums owned by a competitor – is not unique to the Three Tenors. As one PolyGram executive explained: 60 For every major release in any record company there is always an element of anxiety because of big investment, because of big expectations, to make sure that everything is set up to deliver the quantities we need to make money on that project. There was not any difference on this one. JX 97 at 42-43. Recording artists often release material on more than one record label during their careers. Music labels often release an exclusive artist to a competing company for a particular project. Thus, many artists have catalog albums that appear on a label different from the label that releases the artist’s new record. IDF 331-32. In addition, a music label may release an artist from an exclusive recording contract in return for a royalty on the artist’s first album on a new label, giving the companies a shared financial interest in the success of a particular album. IDF 333. In such circumstances, Respondents will likely have the same incentives and opportunity to restrict the pricing or advertising of the artist’s catalog albums that led PolyGram and Warner to enter into the Three Tenors moratorium agreement. Respondent UMG is presently engaged in other joint venture activity – including a joint venture with Sony to distribute music over the Internet – that may provide similar incentives and opportunity to restrain competition. UMG, Sony, and other music companies will provide music to the joint venture on a non- exclusive basis, meaning that music products marketed by the joint venture may also be marketed through traditional retail outlets. Absent a cease and desist order, UMG and Sony may find it profitable to fix prices on product sold to retail stores so as to enhance the joint venture’s sales. IDF 334. We find that, under these circumstances, there is a reasonable risk that Respondents will repeat the unlawful conduct absent an order to cease and desist. See United States v. W.T. Grant Co., 345 U.S. 629, 632 (1953); Marlene’s, Inc. v. Federal Trade Commission, 216 F.2d 556, 560 (7th Cir. 1954); Superior Court Trial Lawyers Ass’n, 107 F.T.C. 510, 602 (1986). 61 V. CONCLUSION At the conclusion of Turandot, the Princess – overcome by the power of love – has a dramatic change of heart. She gladly weds the new suitor and presumably becomes a more kindly ruler. Because we hardly expect those in the business world to act on the basis of such sentiments, we rely on laws and institutions to ensure that businesses adhere to the principles of free competition that keep our economy vigorous and maximize the welfare of consumers. The process of adjudication is vital to those laws, in that it serves to clarify the acceptable bounds of business conduct. In this case, we find that the moratorium agreement between PolyGram and Warner unreasonably restrained trade and constitutes an unfair method of competition. Respondents’ restraints on price discounting and advertising are inherently suspect, because experience and economic learning consistently show that restraints of this sort dampen competition and harm consumers. Respondents’ only proffered justification is not cognizable because it represents a collusive determination that consumers should be deprived of the vigorous competitive offering of certain products to induce them to choose others. Competing busi- nesses contemplating such strategies should be aware that they are antithetical to the fundamental policies of our antitrust laws and will not be countenanced. ISSUED: July 24, 2003