A Primer on Vertical Restraint Law Applied to Intellectual Property – Is Microsoft Really a Vertical Restraint Case?
By Kevin J. O'Connor n1
I. General Concepts
A. First and foremost, the antitrust treatment of intellectual property is not different from other types of property. The D.C. Circuit Court of Appeals succinctly underscored this point in the Microsoft case when analogizing Microsoft's ownership interest in its intellectual property to an ownership interest in a baseball bat.
[Microsoft] claims an absolute and unfettered right to use its intellectual property as it wishes: "[I]f intellectual property rights have been lawfully acquired," it says, then "their subsequent exercise cannot give rise to antitrust liability." Appellant's Opening Br. at 105. That is no more correct than the proposition that use of one's personal property, such as a baseball bat, cannot give rise to tort liability. As the Federal Circuit succinctly stated: "Intellectual property rights do not confer a privilege to violate the antitrust laws."
United States v. Microsoft, 253 F.3d 34, 63 (D.C. Cir. 2001) (en banc) (see Appendix A for detailed history of Microsoft litigation).
B. Because licensing is generally procompetitive, the lawfulness of most licensing arrangements is evaluated under the rule of reason, not the rule of per se illegality. As with other types of property, the per se rule of illegality most often applies to agreements between direct competitors or vertical agreements such as licensing agreements containing an agreement on the minimum level of resale price of the intellectual property or products incorporating the intellectual property.
n1 Kevin J. O'Connor is a shareholder/partner in the law firm of LaFollette Godfrey & Kahn, Madison, WI He is the former Chair of the Multistate Antitrust Task Force of the National Association of Attorneys General.
C. The 1995 Department of Justice Antitrust Division/Federal Trade commission Antitrust Guidelines for the Licensing of Intellectual Property ("IP Guidelines") govern the federal antitrust agencies' application of the antitrust laws to intellectual property. The IP Guidelines are available on the DOJ web site .
D. Recent cases suggest that care should be taken in designing licensing agreements and in settling intellectual property litigation to avoid antitrust problems. This is especially true given that where antitrust problems exist, they can often be avoided and the client's business objectives achieved by modifying the transaction in modest ways.
II. Licensor and Licensee as Direct Competitors
Where the licensor and licensee are direct competitors in a relevant market, any arrangements between them must be scrutinized carefully given the risk that the arrangement might be considered per se illegal. For example, a common problem respecting territorial or field-of-use restrictions occurs where the licensor and licensee are competitors or potential competitors and the agreement effectively divides geographic markets, product markets or customers. Antitrust enforcers often treat such agreements as per se illegal market allocation agreements. See, e.g., United States v. Singer Mfg. Co., 374 U.S. 174 (1963).
III. Resale Price Maintenance
A. Basic Law Is Equivalent To Law Involving Other Products
Setting minimum resale prices in an agreement between licensor and licensee is per se unlawful. This is sometimes referred to as resale or retail price maintenance ("RPM") or vertical price fixing. The decision in State Oil v. Khan, 522 U.S. 3 (1997) established that maximum price fixing will be evaluated under the rule of reason. As is generally the case, it is lawful under the Colgate doctrine for a seller to suggest resale prices on the grounds that such "suggestions" are considered a unilateral action by the seller. Hence, where a seller announces that it will only sell to buyers who sell above a suggested minimum price, the Colgate doctrine will protect the seller from per se liability. Where a seller and a buyer subsequently enter into an agreement to abide by suggested resale prices, or a buyer is coerced or persuaded to comply with the suggested prices, there may be a violation.
B. Evolution of the Law
Early Supreme Court authority provided that patent licenses were exceptions to the general rule against RPM agreements. United States v. General Electric, 272 U.S. 476 (1926). See also Columbia Pictures Corp. v. Coomer, 99 F. Supp. 481 (D. Ky. 1951) (RPM permitted for copyrighted goods although agreement on release prices unlawful). Subsequent cases have held RPM agreements unlawful in virtually every context that has been considered thereby reducing the precedential value of General Electric. See, e.g., United States v. United States Gypsum Co., 333 U.S. 364 (1948); United States v. Univis Lens Co., 316 U.S. 241, 252-53 (1942). The IP Guidelines clearly state that the government will challenge RPM agreements in the intellectual property context. IP Guidelines § 5.2.
C. Practice Pointers
This is an area that requires a great deal of caution given the likely application of the per se rule to any agreement to set or constrain resale prices. Counsel should consider ways to achieve the client's objectives without constraining the pricing freedom of downstream buyers. For example, where a licensor is concerned that its intellectual property may not be resold at a price sufficiently high to generate expected royalties, rather than attempting to set a minimum resale price in the licensing agreement, the royalties could be based on a percentage of the actual resale price or the suggested resale price, whichever is higher. Alternatively, where the intellectual property is being resold with other products, the royalties could be based on a percentage of total sales or IP sales, whichever is greater.
IV. Output Restrictions
A. General Approach
Maximum output restrictions, under standard economic reasoning, serve the same purpose as resale price maintenance agreements, that is, to keep the price of goods higher. Unlike the treatment of resale price maintenance agreements, however, the courts have generally evaluated such output restrictions under the rule of reason and have upheld them at times. See, e.g., Atari Games Corp. v. Nintendo of America, Inc., 897 F.2d 1572, 1578 (Fed. Cir. 1990); Q-Tips, Inc. v. Johnson & Johnson, 109 F. Supp. 657 (D.N.J. 1951), aff'd on other grounds, 206 F.2d 144 (3d Cir. N.J. 1953), cert. denied, 347 U.S. 935 (1953). Although the courts appear to take a lax view of output restrictions on the general theory that the licensor could have simply declined to grant a license (thereby causing zero output), the I.P. Guidelines make it clear that output restrictions may be considered per se unlawful at least where the parties to the agreement are competitors. IP Guidelines § 3.4. Finally, there is some uncertainty as to the treatment of output restrictions that apply to unpatented products produced with a patented process. In Q-Tips such a limitation was upheld but prior cases suggest that such limitations might be viewed as an attempt to extend the patent monopoly beyond the scope of the initial patent thereby constituting misuse.
B. Practice Pointers
Output restrictions should be strongly discouraged where the licensor and licensee are competitors in the relevant market or related markets. If the parties to such an agreement are not competitors, it is important to understand both the technology and the business reasons for the output restrictions. Strong consideration should be given to use of proportionately larger percentage royalties as output reaches certain levels, rather than direct restrictions on output.
V. Exclusive Licenses and Exclusive Dealing Arrangements
A. General Concepts
An exclusive license typically gives the licensee exclusive rights over the intellectual property contained in the license. In contrast, an exclusive dealing arrangement usually contains an agreement by the licensee not to manufacture or sell products made with competing technology. As discussed below, both types of agreements can be extremely problematic where the licensor and licensee are actual or potential competitors. Where the parties are not competitors, however, an exclusive license is generally not problematic under the antitrust laws under the general theory that the licensor could have refused to license at all, hence, granting exclusive rights to intellectual property to a licensee does not on its face reduce competition. On the other hand, exclusive dealing requirements imposed upon licensee can be quite problematic especially where the licensor's intellectual property commands a very high market share, thereby preventing potentially the introduction of new, competitive technologies.
B. Licensor and Licensee as Competitors
Even an exclusive licensing agreement can be problematic under the antitrust laws where the licensor and the licensee are actual or potential competitors. Such exclusive licenses can reduce or eliminate the incentive to compete or even the ability to compete. Where exclusive cross-licenses are involved, these affects can be amplified and the antitrust risk heightened. Additionally, where exclusive licenses incorporate provisions which reduce competition between the licensor and licensee in other ways (e.g., territorial or customary restrictions), the agreement may be viewed as a per se unlawful horizontal agreement between competitors.
Exclusive dealing arrangements may directly constrain the licensee's ability to use competing technologies or may do so indirectly by virtue of incentive provisions which reduce or eliminate the economic incentive to use alternative technology. See e.g., United States v. Microsoft, Inc., Civ. No. 94-1564 (D.D.C., July 15, 1994) (per system royalty on Windows operating system required all personal computers sold by computer manufacturers regardless of the type of operating system sold with the computer). Even where such restrictions do not amount to an antitrust violation, such exclusive dealing restrictions can constitute copyright misuse. See, e.g., Lasercomb America, Inc. v. Reynolds, 911 F.2d 970, 977-79 (4th Cir. 1990).
C. Practice Pointers
Where exclusive licensing or exclusive dealing arrangements are contemplated, antitrust risk is likely to be high where the licensor and licensee are direct competitors. It is very important that counsel seek ways to structure the transaction in a way that will minimize antitrust risk. Where such arrangements are between non-competitors, exclusive licensing agreements can be considered low-risk and exclusive dealing arrangements can be considered moderate to low risk depending upon the likely impact of exclusive dealing arrangement on the licensee's market(s). Where a licensee is required to use the licensor's technology exclusively, the anticompetitive effect of such a restriction is likely to be out weighed by the procompetitive effects if only a small percentage of the licensee's market is impacted and competing technology sellers have alternative, prospective licensees ready and able to enter into licensing agreements. It is important to understand both the business reasons for such exclusive arrangements and the likely method of analysis under the antitrust laws and the I.P. Guidelines.
VI. Patent Infringement-Related Settlements
A. General Concepts
Genuine settlements of potential or actual patent litigation that do not extend the reach of the patent are usually not anti-competitive. But patent infringement-related settlements - particularly in the pharmaceutical context - are carefully scrutinized by agencies and others, and can constitute anti-competitive conduct. Particularly risky are monetary agreements to not market an approved generic drug. Such agreements can have the effect of preventing or delaying other generic manufacturers from entering the market. See, e.g., In the Matter of Abbott Labs., 65 Fed. Reg. 17502 (April 3, 2000); In the Matter of Hoechst Marion Roussel Inc., 66 Fed. Reg. 18636 (April 10, 2001). A generic drug manufacturer's agreement to delay introduction of a product in exchange for payments from the brand name manufacturer has been held to be a per se illegal restraint of trade. See In re: Cardizem CD Antitrust Litigation, 332 F.2d 896 (6th Cir. 2002). A settlement agreement between a brand name manufacturer and generic companies that results in increased competition does not create an antitrust injury, but such a settlement could be deemed part of a larger anti-competitive scheme. See SmithKline Beecham Corp. v. Apotex Corp., 2004 WL 2222388 (E.D. Pa. Sept. 29, 2004).
In determining antitrust liability where a patent-related settlement is at issue, courts focus on: the scope of the patent's exclusionary potential; the extent to which the agreement exceeds that scope; and the arrangement's resulting anti-competitive effects. Schering-Plough Corp. v. FTC, 402 F.3d 1056, 1066 (11th Cir. 2005). Thus, a "reverse payment" from a patent holder to a generic drug company as part of a settlement after the district court ruled against the patent holder, and while the appeal was pending, was held to not render the agreement per se illegal and did not constitute an antitrust violation, absent a showing that its exclusionary effects exceeded the scope of the patent's protection. Joblove v. Barr Labs., Inc. (In re Tamoxifen Citrate Antitrust Litig., 429 F.3d 370 (2d Cir. 2005).
B. Practice Pointers
If an agreement or settlement related to a patent infringement matter is under consideration, the risks in the pharmaceutical context will be high where there are provisions for: reverse payments, that is, payments from patent holders to generic entrants; restrictions on the generic manufacturer's ability to enter the market with noninfringing products; or restrictions on the generic manufacturer's ability to assign or waive its 180-day statutory exclusivity period.
Outside of pharmaceuticals, 180-day exclusivity is not an issue. But proposed settlements that contain provisions for payments to the new entrant or restrictions on the new entrant's ability to enter the market with non-infringing products should be carefully evaluated. Although a settlement can be pro-competitive where it allows a new entrant to market its product sooner than would have been the case had the litigation run its course or the new entrant lost the case, companies should be prepared for considerable scrutiny of a settlement's effects on competition.
VII. Field of Use, Territorial and Customer Restrictions
A. General Concepts
As their name implies, these restrictions limit either the geographic area in which a licensee may sell, the type of products that may be manufactured or sold by using the specific technology, or the types of customers to which a licensee may sell. In general, these restrictions are analyzed under the rule of reason when the licensor and licensee are not actual or potential competitors with one exception. The exception is that territorial restrictions in patent licenses are permitted under Section 261 of the Patent Code, 35 U.S.C. § 261. This exception only applies in territorial restrictions in the United States and only to the first sale of the intellectual property. Generally speaking, the procompetitive and anticompetitive aspects of such restrictions are balanced against each other under the rule of reason. I.P. Guidelines § 2.3. Where a licensor has significant market power, care should be taken to limit territorial restrictions to only the licensor's intellectual property and not to attempt to apply the restriction to the licensee's use of competing technology. See, e.g., United States v. Pilkington plc., Civ. No. 94-345 TUC-WDB (D. Ariz., May 25, 1994).
B. Practice Pointers
First and foremost, it is critical to understand whether or not the licensor and licensee are competitors or potential competitors. If so, these types of restrictions can pose significant antitrust risk. If it appears that the licensor and licensee are competitors, extreme care should be taken to analyze the competitive implications of the agreement and to document scrupulously the business rationale for the restrictions. If the licensor and licensee are not competitors, the antitrust risk is low to moderate. Notwithstanding this, care should be taken to document the business rationale and procompetitive benefits from such agreements as well.
VIII. Royalties, Grantbacks, and Related I.P. Issues
A. General Concepts
Extension of royalties beyond the term of the IP grant, royalties conditioned on the total sales of the licensee, discriminatory royalty, grantbacks required by licensing agreements, and related conditions are often treated under the law of patent and copyright misuse rather than the antitrust law. Hence, an extensive discussion of these principles is not warranted here. However, it should be noted that the IP Guidelines do deal with some aspects of these practices. For example, the IP Guidelines state that non-exclusive grantback agreements will not usually pose competitive problems. However, exclusive grantbacks may prevent antitrust issues where the licensor or licensee possess market power. I.P. Guidelines § 5.6.
"Reach-through royalty" arrangements have been the subject of recent litigation. Typically, these arrangements involve "enabling" technologies that are used as tools in developing commercially viable products. For example, a licensor of a drug screening technology might arrange to be paid not only a royalty for use of the "tool" but also a percentage royalty on sales of the drugs developed with the tool. Although the subject of IP cases, the practice can impact competition if the reach-through royalty burden on sellers of commercial products becomes great enough. Such arrangements have survived recent challenges but the law is still in an embryonic state. See, e.g., Bayer AG v. Housey Pharms. Inc., 169 F. Supp. 2d 328 (D. Del. 2001); Integra Lifesciences I, Ltd. v. Merck KGaA, 331 F.3d 860 (Fed. Cir. 2003).
B. Practice Pointers
It is important to understand the business reasons for varying royalties and grantback provisions. Generally, in the absence of market power, royalty provisions and grantback provisions will likely not be considered to be antitrust issues or even intellectual property issues. In general, the safest course is to insure that royalty provisions do not exceed the life of the intellectual property right and that grantback provisions do not significantly exceed the scope of the original intellectual property right.
A. Basic Concepts
Conditioning the ability of a licensee to license one or more items of intellectual property on the licensee's purchase of another item of intellectual property or a good or a service has been held in some cases to constitute illegal tying. See, e.g., Eastman Kodak Co. v. Image Technical Servs., 112 S. Ct. 2072, 2079 (1992); United States v. Paramount Pictures, Inc., 334 U.S. 131. 156-58 (1948). The IP Guidelines § 5.3 indicate that the federal agencies will apply the same law regarding tying as is applied to other products.
In particular, the licensing of multiple items of intellectual property in a single license or in a group of related licenses, so-called "package licenses" are likely to be treated as tying arrangements. This will result in the efficiencies of the packaging being weighed against any anticompetitive aspects of the package licenses.
In Ill. Tool Works, Inc. v. Indep. Ink, Inc., 126 S.Ct. 1281, 1293 (2006), all eight Supreme Court justices participating in the case concluded that a patent "does not necessarily confer market power upon the patentee." Rejecting the conclusion of the lower court, Independent Ink, Inc. v. Illinois Tool Works, Inc., 396 F.3d 1342, 1348-49 (Fed. Cir. 2005), the Supreme Court held that a plaintiff alleging illegal tying to a patented product must prove that the defendant has market power in a patented tying product, and does not enjoy any presumption of market power.
In another recent tying case, Monsanto Co. v. Scruggs, 459 F.3d 1328, the Federal Circuit upheld Monsanto's downstream restriction preventing farmers from using second generation seeds grown from patented Monsanto seeds, concluding that the patent exhaustion doctrine did not apply. Id. at 1336. The appellants' alleged that Monsanto was illegally tying the purchase of its seed to the purchase of its "Roundup" brand herbicide through license restrictions, incentive agreements, and seed partner agreements. The court rejected this tying claim for lack evidentiary support, noting that Monsanto's grower incentive and seed partner agreements did not force or coerce participants to buy Roundup. The court also concluded that Monsanto did not tie the sale of seeds with "Roundup Ready" genes to seeds with other popular genetic traits, because it offered seeds that did not include both features.
B. Practice Pointers
Where a licensor has market power in a tying product involving intellectual power, care should be taken when bundling IP licenses. Giving prospective licensees the option of licensing the IP associated with market power separately from any package license that might include it can lessen the antitrust risk. It is important to appreciate that the agencies will not presume market power merely because of the presence of IP rights. Hence, where a properly defined market does not indicate the presence of market power in the tying product, there will be negligible antitrust risk under a tying theory.
X. Standard Setting
A. Basic Concepts
Standard setting by direct competitors is often very pro-competitive and efficiency enhancing especially where such standards facilitate the compatibility of substitute and complementary products by competing sellers. Where standard setting organizations are used to exclude competing products from the market or participants in the standard setting process fail to disclose their ownership of blocking IP rights in the emerging standards (i.e., "patent ambushes"), antitrust risk rises. Fairness and equal access are the traditional focus of the federal agencies in this area. See, e.g., Business Review Letter concerning the MPEG-2 Patent Pool (June 26, 1997). More recently, the agencies have been concerned with "patent ambushes." See, e.g., In the Matter of Rambus, Inc., F.T.C. Docket No. 9302 (2002) (complaint); In the Matter of Union Oil Company of California, Docket No. 9305 (2003); In re Dell Computer Corp., 121 F.T.C. 616 (1996).
In Union Oil, the ALJ found that the conduct of Union Oil (commonly known as "Unocal") involving the California Air Resources Board, an agency of the California Environmental Protection Agency, constituted political petitioning activity immune from antitrust liability. F.T.C. Docket No. 9305 (Nov. 25, 2003). The FTC reversed and vacated the ALJ decision holding that a misrepresentation to a government agency outside of the political arena where the misrepresentation is "deliberate, factually verifiable, and central to the outcome of the proceeding or case" is not entitled to Noerr-Pennington protection. F.T.C. Docket No. 9305 (July 6, 2004). The FTC ultimately settled its "patent ambush" case against Unocal when both Unocal and its proposed merger partner, Chevron, agreed to stop all efforts to assert patents on certain low-emission gasoline innovations F.T.C. Docket No. 9305 (August 2, 2005).
In Rambus, FTC regulators charged Rambus with deceiving an industry standard-setting group by participating in cooperative standard-setting activities without disclosing that it was seeking patents on specific technologies ultimately adopted as industry standards. F.T.C. Docket No. 9302 (June 18, 2002). The ALJ found that the standard-setting organization at issue did not have sufficiently clear rules on disclosure to impose a disclosure duty. F.T.C. Docket No. 9302 (Feb. 23, 2004). The ALJ further found that a failure to comply with a standard setting organization's disclosure rules is not, in and of itself, a sufficient basis for antitrust liability.
In August 2006, however, all five Federal Trade Commissioners issued a ruling reversing this initial decision on the merits of the original complaint. F.T.C. Docket No. 9302 (August 2, 2006). The August 2006 FTC opinion cited "new proposed findings of fact and conclusions of law that were unavailable to the ALJ" in concluding that Rambus had indeed violated Section 5 of the FTC Act. Id., Opinion of the Commission at 4. Rambus had run afoul of the Act, the FTC concluded, by engaging in exclusionary conduct such as failing to "disclose the existence of its patents and applications" for technology being considered by the standard-setting group, taking "additional actions that misled" the group, and by "amend[ing] its patent applications to ensure that subsequently-issued patents would cover the ultimate standard." Id.
The commission held oral arguments on the question of remedies in November 2006.
C. Practice Pointers
Participants in standard setting organizations need to be sensitive to the traditional concerns of unfair exclusion of rivals from markets impacted by the standard setting activity. More recently, holders of IP rights implicated in the standards, need to make careful choices early in the process concerning disclosure of patent rights that may underpin particular standards under consideration. This area of law continues to evolve.
XI. Top Ten Antitrust Risks in IP Licensing
10. An exclusive license forecloses the ability of the licensee to use competing technologies.
9. The licensor engages in "patent ambush" during a standard setting process.
8. The license increases the costs of competitors of obtaining important inputs to their business.
7. The licensor and licensee cannot demonstrate any legitimate business reason for the restrictions in the license.
6. The royalties are structured to impose effectively an exclusive dealing arrangement especially where the licensor has market power.
5. The licensor attempts to profit from the sale of products made from the licensor's technology to such an extent that competition is affected downstream.
4. The licensor attempts to tie the use of its technology to the purchase of other products.
3. The licensor engages in resale price maintenance of the reselling of the technology or products manufactured with the technology.
2. The license is between direct competitors and restricts output or allocates customers.
The license is between direct competitors and constrains pricing freedom: i.e., price fixing or bid rigging.
IP ANTITRUST ISSUES IN THE MICROSOFT CASE
Is Microsoft Really a Distribution Restraints Case?
I. Procedural History of the Microsoft Case.
A. The United States (Antitrust Division) filed a civil complaint on May 18, 1998, alleging that Microsoft had engaged in anticompetitive conduct in violation of Sections 1 and 2 of the Sherman Act, 15 U.S.C. §§ 1, 2. Simultaneously, twenty states and the District of Columbia (hereinafter "the States") filed a similar, but not identical, action challenging much of the same conduct. At Microsoft's request, the two cases were consolidated. (The cases were de-consolidated on February 2, 2002.)
B. The United States and the States coordinated the pretrial preparation and trial of the case. In a bench trial lasting from October 19, 1998, to June 24, 1999, (78 trial days), the court heard testimony from 26 witnesses, admitted depositions of 79 other witnesses and admitted 2733 exhibits. On November 5, 1999, the court entered 412 findings of fact. United States v. Microsoft Corp., 84 F. Supp. 2d 9 (D.D.C. 1999) ("Findings of Fact").
C. Between November 1999 and April 2000, the parties attempted unsuccessfully to settle the suit through mediation before antitrust scholar Judge Richard Posner.
D. The district court then entered its conclusions of law on April 3, 2000. United States v. Microsoft Corp., 87 F. Supp. 2d 30 (D.D.C. 2000) ("Conclusions of Law"). After further proceedings on remedy, the district court entered its final judgment, which judgment included provision for the breakup of Microsoft and interim conduct relief. 97 F. Supp. 2d 59 (D.D.C. 2000) ("Initial Final Judgment" (IFJ)).
E. The district court ruled that Microsoft successfully had engaged in a series of anticompetitive acts to protect and maintain that monopoly, in violation of Section 2 of the Sherman Act. Conclusions of Law at 37-44. (Note: the plaintiffs never contended that Microsoft unlawfully had obtained its monopoly in Intel-compatible personal computer (PC) operating systems.). The court also ruled that Microsoft unlawfully had attempted to monopolize the Internet Web browser market and had tied its Web browser, Internet Explorer (IE), to its Windows operating system. Id. at 45-51. The district court rejected plaintiffs' claim that Microsoft's exclusive dealing contracts violated Section 1 of the Sherman Act. Id. at 51-54.
F. To remedy the violations, the court ordered Microsoft to break up into separate operating system and applications businesses. Initial Final Judgment, 97 F. Supp. 2d at 64-65. The Initial Final Judgment also ordered interim conduct restrictions until the structural relief became effective. Id. at 66-69.
G. Microsoft filed notices of appeal, and the Court of Appeals, sua sponte, ordered that any proceedings before it be heard en banc. The district court certified the case for direct appeal to the Supreme Court pursuant to the Expediting Act of 1903, as amended, 15 U.S.C. § 29(b), and stayed its judgment pending completion of the appellate process. Order (June 20, 2000). The States filed a petition for certiorari as well. The Supreme Court declined to accept the appeal, denied the States' cert. petition, and remanded the case to the Court of Appeals. Microsoft Corp. v. United States, 530 U.S. 1301 (2000).
II. Court of Appeals Decision.
A. After extensive briefing and two days of oral argument, the en banc Court of Appeals issued a unanimous (7-0) and comprehensive decision affirming in part, reversing in part, and remanding in part for proceedings before a different district judge. United States v. Microsoft Corp., 253 F.3d 34 (D.C. Cir. 2001) (en banc) ("Microsoft").
B. The Court of Appeals affirmed the district court's principal ruling that Microsoft unlawfully had maintained its operating system monopoly by a series of acts that eliminated two "middleware" threats that could have eroded Windows favorable position. Id. at 50-80. "Middleware" is platform software that runs on an operating system but also exposes its own application programming interfaces (APIs) so that other applications can run on the middleware itself. Id. at 53; Findings of Fact, P 28. Where the middleware is cross-platform (i.e., able to run on multiple operating systems), an applications program written to rely exclusively on a middleware program's APIs could run on a computer regardless of the operating system running. Thus, the Court of Appeals found that middleware poses a potential threat to Microsoft's operating system monopoly because if enough applications developers (known as independent software vendors (ISVs)) write enough applications for widely used middleware, computer users no longer would be reluctant to choose a non-Windows operating system for fear that it would run an insufficient array of applications. Id. at 53. Over time, this dynamic would erode the "applications barrier to entry" that protected Microsoft's Windows monopoly.
C. Microsoft's anticompetitive acts centered on two middleware threats: Netscape's web browser (Navigator), and Sun Microsystem's Java technologies. To eliminate Navigator's threat, Microsoft set out to ensure that its own web browser, IE, gained dominant usage so that ISVs would continue to focus their efforts on the Windows platform rather than the Navigator platform. Microsoft took steps to constrict Netscape's access to the distribution channels.
D. Java technologies enabled developers to write programs that could be ported to different operating systems with relative ease thereby posing a middleware threat to Microsoft. In 1995, Netscape had announced that it would include a Java Virtual Machine (JVM) with every copy of Navigator, thereby creating the possibility that Sun's Java implementation would achieve the necessary ubiquity on Windows to pose a threat to the applications barrier to entry. Id. at 74. Thus, by limiting the usage of Navigator, Microsoft simultaneously limited the distribution of Java. In addition, Microsoft took additional steps to inhibit the distribution of Java: (1) pressuring developers not to support cross-platform Java (id. at 75); (2) seeking to eliminate the Java threat through technological means that maximized the difficulty with which applications written in Java could be ported from Windows to other platforms, and vice versa (id. at 74-75); and (3) other anticompetitive steps to discourage developers from creating Java applications compatible with non-Microsoft JVMs (id. at 75-77).
E. In affirming liability for monopoly maintenance, however, the Court of Appeals upheld twelve of the twenty district court findings that particular acts constituted bases for violations of Section 2. See id. at 59-78. These findings can be grouped into four categories of conduct including that Microsoft:
1. undertook a variety of restrictions on OEMs (253 F.3d at 59-64);
2. integrated its Web browser, IE, into Windows in a non-removable way while excluding rivals (id. at 64-67);
3. engaged in restrictive and exclusionary dealings with IAPs, ISVs, and Apple Computer (id. at 67-74); and
4. threatened and attempted to mislead software developers in order to contain and subvert Java middleware technologies that threatened Microsoft's operating system monopoly (id. at 75-77).
F. The court rejected certain trial court findings that Microsoft had violated Section 2 by:
1. prohibiting OEMs from "automatically launching a substitute user interface upon completion of the boot process" (id. at 63);
2. overriding the user's choice of browser in certain circumstances (id. at 67);
3. giving away its Internet Explorer browser to IAPs and ISVs (id. at 67-68, 75);
4. offering IAPs a bounty for each customer the IAP signs up for service using the IE browser (id. at 67-68);
5. developing and giving away the Internet Explorer Access Kit (IEAK) (id. at 68);
6. entering into exclusive agreements with Internet Content Providers (ICPs) (id. at 72);
7. creating a JVM that runs faster on Windows but lacks the cross-platform attributes that Sun's (hence Navigator's) JVM possesses (id. at 74-75).
G. The Court of Appeals expressly rejected the district court's conclusion that, "apart from Microsoft's specific acts, Microsoft was liable under § 2 based upon its general 'course of conduct.'" Id. at 78. The court found that the district court had failed to "point to any series of acts, each of which harms competition only slightly but the cumulative effect of which is significant enough to form an independent basis for liability." Id.
H. The Court of Appeals also reversed the district court's conclusion regarding Microsoft's attempt to monopolize the Web browser market in violation of Section 2. Id. at 80-84. The court found that plaintiffs had failed to define and prove a market for Web browsers, a necessary element of the claim. Id. at 81-82.
I. The Court of Appeals vacated the district court's judgment on the Section 1 tying claim as well, but remanded that claim to the district court for reconsideration under the rule of reason. Id. at 84-97. In dicta, the Court of Appeals held that the market for platform software presented unique issues under tying law suggesting a balancing test in place of the per se rule. The court directed that on remand, plaintiffs would be limited to proving that the anticompetitive effects from tying outweigh the benefits in the tied product market, not just that those effects outweigh the benefits overall. Id. at 95. In addition, plaintiffs would be "precluded from arguing any theory of harm that depends on a precise definition of browsers or barriers to entry . . . other than what may be implicit in Microsoft's tying arrangement." Id.
J. In light of its determination that it had "drastically" (id. at 105, 107) altered the district court's conclusions on liability, and its finding that an evidentiary hearing on remedy was necessary (id. at 101-103), the Court of Appeals vacated the final judgment and remanded the case to the district court for further proceedings. Id. at 107. The court also offered "guidance . . . to advance the ultimate resolution of this important controversy" suggesting that a structural remedy was not supported by the evidence. Id. at 105.
K. The Court of Appeals admonished the district court on remand to bear in mind the role of causation when fashioning relief, directing this Court to "consider whether plaintiffs have established a sufficient causal connection between Microsoft's anticompetitive conduct and its dominant position in the [operating system] market." Id. at 106. Absent "clear" indication of a "'significant causal connection between the conduct and creation or maintenance of the market power,'" Microsoft's unlawful behavior "should be remedied by 'an injunction against continuation of that conduct.'" Id. at 106 (quoting 3 Phillip E. Areeda & Herbert Hovenkamp, Antitrust Law P 650a, at 67 (rev. ed. 1996)) (emphasis added by Court of Appeals).
L. Finally, the Court of Appeals concluded that the district judge's contacts with the press violated the Code of Conduct for United States Judges and warranted disqualification under 28 U.S.C. § 455(a). Id. at 107-118. The court vacated the remedy on the additional basis that the district judge's misconduct infected the remedial phase. Id. at 117. The Court of Appeals rejected Microsoft's procedural challenges to the trial court proceedings, finding the district court's actions "comfortably within the bounds of its broad discretion to conduct trials as it sees fit." 253 F.3d at 98, 100-01.
III. Remand to District Court.
A. After the decision, the Court of Appeals rejected Microsoft's petition for rehearing which had attempted to revise the decision to reverse the trial court's finding that Microsoft's intermingling of code between the browser and the operating system was anticompetitive. Microsoft filed a petition for a writ of certiorari based on the Court of Appeals' failure to vacate the Findings of Fact and Conclusions of Law -- and not just the remedy -- in light of the district judge's misconduct. Petition for a Writ of Certiorari, No. 01-236 ("Cert. Petition") was denied. Microsoft Corp. v. United States, 122 S. Ct. 350 (2001).
B. On September 28, 2001, the new trial judge ordered the parties into a new round of intense settlement negotiations and probable mediation. The Court directed plaintiffs to "determine which portions of the former judgment remain appropriate in light of the appellate court's ruling and which portions are unsupported following the appellate court's narrowing of liability." Tr. 9/28/01 at 8. The Court also adopted a fast-track discovery and evidentiary hearing schedule in case the parties failed to settle.
C. On November 2, 2001, following five weeks of intensive negotiation and mediation, the United States and Microsoft agreed on terms of a proposed final judgment. Further negotiations with several of the plaintiff States resulted in submission on November 6, 2001, by the United States, the Settling States (New York, Ohio, Illinois, Kentucky, Louisiana, Maryland, Michigan, North Carolina, and Wisconsin), and Microsoft of the Revised Proposed Final Judgment (RPFJ). As of April 2002, the RPFJ is pending before the district court. The RPFJ provides for OEM protection, nondiscriminatory treatment of ISVs, the modularization of platform software sold with Windows, disclosure of APIs and other interfaces sufficient to enable the development of compatible applications software and an extensive compliance regime.
D. Several states (California, Massachusetts, Kansas, Utah, Iowa, Connecticut, West Virginia, Minnesota, Florida) and the District of Columbia chose not to enter into the RPFJ. On March 18, 2002, a hearing commenced in the district court on the non-settling states' proposed remedy which includes relief provisions similar to the RPFJ in some respects but also contains other provisions requiring Microsoft to disclose and license the source code for IE, to distribute Java with Windows for ten years, to allow for the porting of Office to other operating systems, and to adhere to industry standards if Microsoft claims that its products are compliant.
E. On November 1, 2002, the District Court issued a ruling essentially approving the settlement between Microsoft, the United States and the settling states with a number of modifications and rejecting most of the proposed relief measures requested by the litigating states. United States v. Microsoft, Inc., 2002 WL 3149450, 2002-2 Trade Cases P 73,851 (D.D.C. Nov. 1, 2002).
IV. Intellectual Property Aspects of the Microsoft Litigation.
A. At various points during the litigation, Microsoft attempted to raise defenses to the core antitrust allegations based not on antitrust law but on copyright law. Microsoft's motion for summary judgment on the grounds that its IP rights shielded its conduct was denied in a decision on September 14, 1998. 1998 WL 614485 at *15 (D.D.C. Sept. 14, 1998).
B. Even though Microsoft had been put on notice that a blanket exemption was not available under the copyright law, Microsoft did not offer any evidence at trial related to the IP defense issue other than the fact that Microsoft's operating systems received copyright protection, a fact not disputed by the government. As a result, the trial judge rejected the IP defense stating that it is "well settled that a copyright holder is not by reason thereof entitled to employ the prerequisites in ways that directly threatens competition." United States v. Microsoft Corp., 87 F. Supp. 30, 40 (D.D.C. 2000). Indeed, given the lack of vigor with which Microsoft pursued this defense at trial, the trial began his discussion of the IP issues stating "To the extent that Microsoft still asserts a copyright defense . . ." Id.
C. The district issued only one Finding of Fact dealing directly with the IP issues. Finding 228 stated in its entirety:
Like most other software products, Windows 95 and Windows 98 are covered by copyright registrations. Since they are copyrighted, Microsoft distributes these products to OEMs pursuant to license agreements. By early 1998, Microsoft had made these licenses conditional on OEMs' compliance with the restrictions described above. Notwithstanding the formal inclusion of these restrictions in the license agreements, the removal of the Internet Explorer icon and the promotions of Navigator in the boot sequence would not have compromised Microsoft's creative expression or interfered with its ability to reap the legitimate value of its ingenuity and investment in developing Windows. More generally, the contemporaneous Microsoft documents reflect concern with the promotion of Navigator rather than the infringement of a copyright.
D. Microsoft essentially repeated to the Court of Appeals the arguments that had been made in the trial court, essentially that "the license restrictions are legally justified because, in imposing them, Microsoft is simply 'exercising its rights as the holder of valid copyrights.'" 253 F.3d at 62.
E. The Court of Appeals dealt with Microsoft's IP arguments almost as summarily as the trial court:
Microsoft's primary copyright argument borders upon the frivolous. The company claims an absolute and unfettered right to use its intellectual property as it wishes: "[I]f intellectual property rights have been lawfully acquired," it says, then "their subsequent exercise cannot give rise to antitrust liability." Appellant's Opening Br. at 105. That is no more correct than the proposition that use of one's personal property, such as a baseball bat, cannot give rise to tort liability. As the Federal Circuit succinctly stated: "Intellectual property rights do not confer a privilege to violate the antitrust laws." 253 F.3d at 63 (relevant pages attached).
F. In addition, Microsoft made two arguments to the effect that it was not unreasonably exercising its copyright in an unreasonable manner.
1. First, Microsoft cites Gilliam v. American Broadcasting Cos., 538 F.2d 14 (2d Cir. 1976); WGN Cont'l Broad. Co. v. United Video, Inc., 693 F.2d 622 (7th Cir. 1982) for the proposition that a copyright holder may limit a licensee's ability to engage in "significant and deleterious alterations" of a copyrighted work. The Court of Appeals distinguished those cases from the Microsoft fact pattern on the grounds that those cases involved "substantial alterations" and neither involved any claim that the copyright holder had violated the antitrust laws. 253 F.3d at 63. The only license restriction Microsoft "seriously defends as necessary to prevent a 'substantial alteration' of its copyrighted work is the prohibition on OEMs automatically launching a substitute user interface upon completion of the boot process." Id. The Court agreed that allowing OEMs to load a shell program that prevents the user from even seeing the Windows desktop briefly is a substantial alteration and Microsoft's interest in preventing this outweighs the anticompetitive effect it might have.
2. The second variation on the copyright defense posited by Microsoft is that the license restrictions imposed by Microsoft are necessary to prevent OEMs from so altering Windows as to undermine the principal value of Windows as a stable platform for software development. Id. The only evidence Microsoft submitted at trial relating indirectly to this justification was an exhibit that mentioned the possible consumer confusion that might result if OEMs added programs to the desktop. But the Court rejected this outright noting that the OEMs bear the bulk of the costs caused by consumer confusion in the form of increased consumer complaints and calls for assistance. The only rational explanation for this restriction was that addition of such programs to the desktop undermines Microsoft's monopoly and "that is not a permissible justification for the license restrictions." Id. at 64.
G. Microsoft argued that its restrictions and other actions did not completely bar Netscape from distributing its product. Although noting that this was technically correct, the Court pointed out that Microsoft "did bar [Netscape] from the cost-efficient [means of distribution]." Id. at 64.
H. Microsoft also made arguments justifying its technological bundling of its software with Windows. These attempts to justify its conduct were not explicitly premised on intellectual property rights. However, they provide a useful backdrop to these IP issues in future cases. The three specific circumstances include:
1. In Windows 98, unlike Windows 95, Microsoft did not allow Internet Explorer to be removed using Windows Add/Remove function;
2. Microsoft designed Windows to override the user's choice of a default browser for a variety of Windows functions; and
3. Microsoft had commingled the code from IE and Windows such that any attempt to remove IE files would cripple parts of Windows.
I. First, the Court of Appeals found that each of these restrictions was anticompetitive because it reduced user choice. The second step of the analysis weighed the anticompetitive effect against Microsoft's proffered justifications. With respect to the first and third items, Microsoft's justifications were insufficient. With respect to the aspect of Windows that allows for the override of the user's choice of browser in certain situations, the Court held that such a justification did outweigh the anticompetitive effect.