DEPARTMENT OF JUSTICE

There is, in the United States today, increasing debate over the multi-faceted nature of our antitrust enforcement regime and, in particular, the issues it raises with respect to securing appropriate and effective relief for antitrust violations. On the civil side alone, antitrust violators in the United States potentially confront: (1) two federal agencies, the Department of Justice Antitrust Division and the Federal Trade Commission, each with largely overlapping jurisdiction for civil antitrust enforcement, but with different procedures; (2) additional federal agencies, such as the Federal Communications Commission and the Department of Transportation, that have some specific statutory mandates and authority to consider competition issues in particular matters; (3) 56 "state" attorneys general (including the District of Columbia and five territories) authorized to bring actions under state and federal law; and (4) an infinite number of "private" attorneys general in private class action and individual lawsuits brought by direct purchasers and indirect purchasers down to the consumer level. And, it likely is no coincidence that the debate generated by the multiplicity of actions attacking the same conduct in the United States grows as the number of competition enforcers in the world steadily increases.

With respect to the vitamins cartel, for example, corporate defendants were criminally prosecuted by the Antitrust Division and paid record fines, while certain individual defendants paid fines and served prison sentences. Following on the criminal prosecutions, the defendants were sued by direct purchasers in class actions, by direct purchasers in individual actions, by indirect purchasers in class actions, by indirect purchasers in individual actions, and by states in parens patriae and other actions. Those were just in the United States. The vitamin cartel members have also faced enforcement actions in Canada, the European Union and several other jurisdictions, as well as private lawsuits in some of those countries. While no one would seriously argue that the cartel conduct required stringent relief, even this case has caused many to debate the effectiveness and efficiency of the multi-tiered system.

The United States Department of Justice Antitrust Division has been and remains committed to strong antitrust enforcement and relief as an economic and social imperative. Under-enforcement or failure to effectively remedy a violation harms consumers through higher prices, decreased quality, and reduced output and innovation. Markets are distorted, victims go uncompensated, and violators reap windfalls. But it would be a mistake to assume that for any particular violation "piling on" more enforcement and more relief is always better -- that somehow taking redundant cracks at remedying an antitrust violation automatically results in a stronger enforcement scheme and, ultimately, in stronger competition.

The greatest danger in over-enforcement of competition laws and over-remedying is that they will retard legitimate, pro-competitive behavior -- that is, reduce the very favored conduct that the laws are intended to encourage and protect. To avoid overwhelming costs and burdens, firms may steer so far clear of potentially questionable conduct that they end up avoiding legitimate behavior that may have benefitted consumers. Piling on multiple layers of enforcement or relief may also provide windfalls to other market participants, which further distorts the market away from what competition itself would create. And, of course, it places unnecessary burdens on enforcement agencies, courts, and parties -- costs that ultimately the taxpayers we serve will bear.

As Assistant Attorney General Charles James has stated, layered enforcement should be complementary, not conflicting.⁽¹⁾ Yet, with various enforcers pursuing different concerns and different relief, the potential for conflicting or inconsistent results looms. This means that one enforcer can, with respect to a particular matter or set of matters, set the agenda for all enforcers. As U.S. Federal Trade Commission Chairman Tim Muris said last year with regard to merger review, "The ruling of the most restrictive jurisdiction with respect to a proposed merger ultimately will prevail. Consequently, agreements among regulators may lead businesses to restrict their merger activity to transactions that will be acceptable to all jurisdictions. As a result, merger activity will fall to sub-optimal levels . . . ."⁽²⁾

The dangers of inconsistency and over-enforcement are perhaps most acute in the area of remedies. Once an enforcer secures relief for a particular violation in a particular market, every party with a connection to that market will be affected. The Antitrust Division plays a critical role within the domestic and global patchwork of competition enforcement and takes seriously its responsibility to remedy violations in an effective and responsible manner. We work closely with the Federal Trade Commission, with other federal agencies, with the state attorneys general, and with non-U.S. competition enforcers, in an effort to keep the multi-layered system in balance. We do so not only on a case-by-case basis to avoid particular inconsistencies, but also at the broader policy level with the hope of minimizing conflict going forward. And, we work within our own matters and policy projects to set an example of sound and effective remedies practice. In what follows, I will discuss the Division's policies and actions with respect to remedies through the lens of one policy initiative, our current efforts to develop an internal merger remedies manual, and one major enforcement action, our settlement of United States v. Microsoft Corp.⁽³⁾

During the mid- to late 1990's, the topic of "remedies" in U.S. merger enforcement was widely discussed, as members of the bar and the business community recognized that the vast majority of merger challenges were not litigated, but rather were resolved through negotiation culminating in consent decrees. This reality necessarily focused more attention on the remedy, rather than the violation, because parties to consent decrees do not contest the violation. The debate extended into analyses of what types of remedies are appropriate for settling merger challenges. Interested parties, and particularly counsel representing merging parties, have struggled to discern the agencies' policies on remedies, if any, apart from general practice. Some believe that one negative consequence to "litigation by consent decree" is the lack of global transparency as compared to litigation in court. Some express concern about whether enough remedies have been litigated so that the agencies have appropriate guidance from courts. The FTC's "Study of the Commission's Divestiture Process,"⁽⁴⁾ released in 1999, created further discussion and focused us on the fact that, even after 25 years of merger activity under the Hart-Scott-Rodino model, there is a lot we do not know about remedies.

The Antitrust Division has begun a top-to-bottom review of its practices and policies with respect to seeking and securing remedies in merger enforcement, whether in litigation or through consent decree. Our desire is to provide better guidance, both within and outside of the Division. The FTC is undertaking a parallel review of its remedies practices, and last March, it announced plans to conduct public workshops to explore modifications and improvements to the Commission's use of specific remedy provisions. The two agencies are working together to identify "best practices" in developing and implementing remedies.

We begin our review of remedies from a perspective that includes some important principles. First, the Division will not consider a proposed remedy unless it is confident that there is a violation requiring redress. Parties often propose remedies prior to the conclusion of the Division's investigation, which can efficiently save the taxpayers, the parties, and third parties time and expense. But the Division will not accept remedies simply to avoid investigative work. Consumers would not be benefitted if we secured a "scalp" when there was no violation. Indeed, if we do, they could, in fact, be harmed. Now, obviously, some of the inherent benefits of settlement would be lost if the Division insisted on investigating every merger up to the point of filing litigation before it would consider a resolution. When a party wishes to discuss a possible settlement, the Division will determine whether it has enough information at that point to effectively consider whether there is, in fact, a violation to redress. If the Division determines that it does not yet have enough information but still believes it is in the public interest to discuss or pursue a resolution, the Division will work closely with the parties to obtain quickly the additional information it needs. At bottom, we will not accept remedies for which we would not be prepared to litigate.

Second, the Division must be satisfied that, as far as our predictive abilities will take us, the remedy will effectively redress the violation. Naturally, when parties propose remedies they are trying to guess how much it will take to satisfy the Division so that they can close their deal and move on. As in any other bargaining situation, they tend to start low, wanting to give the bare minimum that is necessary, and they prefer the fix that is least likely to provide them with viable competition going forward. Third parties, such as competitors and customers, when offering remedial suggestions, naturally want a "fix" that is broad or narrow enough to best position them in the marketplace. We agree with the goal of requiring the minimum necessary to maintain competition at premerger levels; the trick is determining what that means. Again, we will not take a likely ineffective fix just to bolster our statistics or to get on to the next matter.

Third, the Antitrust Division is an enforcement agency, not a regulatory body. Consistent with this enforcement role, the Sherman and Clayton Acts authorize the Division to seek and procure a remedy any time it proves a violation of federal antitrust law. But having proven a violation, the goal is not to review the market and decide how it would best operate. Rather, the goal is to effectively remedy the violation for the benefit of consumers, maintaining competition at premerger levels. Once the violation is remedied, competition will decide how the market performs, including choosing the winners and losers.

Fourth, particularly in today's economy, there is no "one-size-fits-all" for remedies. Rather, the Division must be flexible and creative in devising remedies. If remedies do not keep pace with the ever-changing dynamics of competition, they either will not work or will harm the very competition they are intended to maintain. This is not to say that the Division will not continue to consider past precedent where it has worked; our staff's learning on the subject, gained over decades, is invaluable when engaging in the process of choosing an effective remedy. It is true that hard-and-fast rules about what the Division requires upon a determination that a merger would violate Section 7 would make things easier for the Division, as well as for businesses that may crave certainty above all else. We believe, however, that the dangers of tilting remedies in favor of the one-size-fits-all variety outweigh the benefits of the certainty that it brings. It hardly makes sense to undertake a heavily fact-oriented analysis to predict whether a violation has or will occur, only to then use strict rules for devising remedies, without that same fact-based analysis of the likely efficacy.

Fifth, in devising remedies, the Division must be mindful that the remedy may have unintended consequences in the marketplace. Our ability to predict such consequences is, at best, equal to our ability to predict that a merger may substantially lessen competition and that the remedy in the first instance will prevent that lessening. Still, it is something that we must keep in mind given our ultimate goal of protecting competitive markets for the benefit of consumers.

Finally, merger remedies must not be punitive. The only objective in imposing a merger remedy is maintenance of competition at premerger levels. Remedies that extract more than necessary to achieve that goal as a price for permitting the merger to proceed have no place in responsible merger enforcement.

In our internal review of merger remedy policies, we are, for example, exploring when it is appropriate and effective to permit parties to "fix it first," by immediately divesting assets or otherwise changing the terms of the deal without an enforcement action, as contrasted with requiring an "upfront buyer" as part of an enforcement action before the deal may proceed, as the FTC has often done. We are looking at whether "crown jewel" terms, which require parties to sell more desirable assets if they are unable to sell the primary divestiture assets, are ever appropriate. We are debating the efficacy of firewalls; the conditions under which we would accept remedies such as licenses or supply contracts rather than insisting on the divestiture of hard assets; and whether in divestitures we should require "clean sweeps" of one party's ongoing business, rather than allowing a "mix and match" of both parties' assets.

The Microsoft Remedy

To say that the United States' settlement of its antitrust case against Microsoft has prompted extensive debate would be an understatement. In fact, the Microsoft case has long prompted extensive discussion about appropriate remedies in anticompetitive conduct cases, beginning when the Division first sought a break-up of the company and continuing today as the Division awaits approval of its settlement, and as nine states and the District of Columbia hope for further relief. The Division is satisfied that we have obtained the right remedy for the public because we believe our remedy fully resolves the monopoly maintenance claim that the Court of Appeals sustained.

When the case emerged from the Court of Appeals in late June 2001,⁽⁵⁾ we carefully reviewed the opinion and immediately began discussing what it meant for our pursuit of a remedy. Third parties, some of which had been witnesses for the Government in the liability phase, lobbied heavily. In proposing an appropriate remedy, we had no mandate to regulate or to make policy for the computer software industry. Rather, our mandate was to seek to redress the serious antitrust violation for which the Court of Appeals sustained liability and to ensure that the violation would not recur. It was also our strong view that the public interest required redressing the violation as soon as possible.

To place this in context, it is important to trace the case from the beginning. The United States' case originally included four counts: (1) illegal maintenance of Microsoft's monopoly position in operating systems for Intel-based personal computers; (2) attempted monopolization of the web browser market; (3) illegal tying of the web browser, Internet Explorer, to the operating system; and (4) exclusive dealing. The states' case included the same four counts, as well as two additional counts: (1) maintenance of Microsoft's monopoly position in the market for office productivity suite software; and (2) monopoly leveraging.⁽⁶⁾

The states dropped their monopolization claim involving office productivity suite software. The district court granted summary judgment in favor of Microsoft on the states' monopoly leveraging claim.⁽⁷⁾ After trial, the district court found in favor of the United States and the states on monopoly maintenance, attempted monopolization, and tying, but not on exclusive dealing.⁽⁸⁾ The court then imposed the remedy that we and the states had requested, which included interim conduct relief for one year, followed by a break-up of the company into two separate entities.⁽⁹⁾

On appeal, the D.C. Circuit upheld the monopoly maintenance finding based on 12 anticompetitive acts, but rejected eight other specific acts and Microsoft's "course of conduct" as bases for liability.⁽¹⁰⁾ The Court also reversed the attempted monopolization finding of liability; vacated the per se tying finding of liability and remanded the claim for review under the rule of reason standard; and vacated the remedy, providing guidance for the remedies proceeding on remand. In its own words, the court "drastically altered the District Court's conclusions on liability."⁽¹¹⁾ As responsible government enforcers, we had no choice but to account for those alterations as we determined the appropriate remedy to seek. And, indeed, the district court judge stated that she expected us to do so.

The Court of Appeals admonished that a structural remedy would be appropriate only if, on remand, we proved a more direct causal connection between Microsoft's exclusionary practices and maintenance of the operating system monopoly, a connection that the district court had said it could not find. It was clear that the Court of Appeals did not favor a structural remedy. Given the Court's statements on remedy, and the nature of the exclusionary practices that the Court upheld as unlawful, a conduct remedy seemed all that could be secured, let alone justified. Once we came to that conclusion, we immediately informed Microsoft and the court, fearing that holding out the prospect of structural relief as a bargaining chip would only cause delay as Microsoft were permitted to raise and explore a host of issues relating to such a remedy, and potentially would try the patience of the new district court judge, who undoubtedly had read the Court of Appeals' not-so-subtle message about structural relief.

Turning to conduct relief, the remedy had to stop the offending conduct, prevent its recurrence, and restore competitive conditions. It had to go beyond just prohibiting the exact conduct found unlawful to address conduct of a similar nature and to require restorative actions, while still keeping a foundation in the offending conduct. It had to take into account changes in the marketplace. And it had to be implemented as quickly as possible. While we were internally and with the states discussing the appropriate remedies to seek, the district court ordered us to try to settle the case, negotiating 24/7 over a period of five weeks, first without a mediator and, if necessary after a period of time, with a mediator.⁽¹²⁾

Negotiate we did for five weeks, and ultimately we reached a settlement.⁽¹³⁾ Nine states also agreed to settle. Nine other states, and the District of Columbia, however, did not settle and instead went to trial last spring. During the 60-day comment period provided for by the Tunney Act, we received over 30,000 comments, many of them via email. We carefully reviewed all of the comments, made some minor clarifying changes to the settlement agreement, and filed our detailed response in late February.⁽¹⁴⁾ The district court judge held a hearing on the settlement on March 6, 2002 and took the matter under advisement. On July 2, 2002, the court issued an opinion in which she rejected all challenges to the Department's compliance with the Tunney Act and ruled that the Department (and Microsoft) had fulfilled all of its obligations under the Act.⁽¹⁵⁾ Having ruled on those threshold issues, the district court declared that it is now in a position to rule on whether the settlement is in the public interest, and we await that decision.

The settlement we reached with Microsoft stops the conduct found unlawful by the Court of Appeals by broadly banning exclusive dealing, prohibiting a broad range of retaliatory conduct, and giving computer manufacturers extensive control of the desktop. In crafting a remedy that would also prevent the recurrence of the unlawful conduct, we considered the wide range of strategies at Microsoft's disposal, given its position in the market. The settlement prevents recurrence of such conduct using broad concepts of non-discrimination and non-retaliation. In fashioning this remedy, we did, however, recognize that some forms of collaboration are procompetitive, and drafted specific, limited provisions to allow this type of conduct. This is what some firms -- primarily Microsoft's competitors -- have labeled as "loopholes" in the settlement, which I will discuss below. The consent decree also uses a broad, and necessarily complex, definition of middleware. It includes not only current middleware products, such as browsers, email client software, networked audio/video client software, and instant messaging, but also future products that have the potential to be a similar threat to the operating system.

The settlement also restores the lost competitive conditions in the market by restoring the potential threat posed by middleware. Many have argued that the remedy does not go far enough because it does not take away Microsoft's operating system monopoly. However, neither we nor the states alleged that Microsoft unlawfully acquired its monopoly, and the district court and Court of Appeals found that middleware represented a potential or nascent threat -- not necessarily a current threat that would topple the operating system monopoly. The consent decree restores the potential threat through a series of disclosure and design provisions. Microsoft must disclose all application programming interfaces ("APIs") in the operating system if Microsoft middleware products rely on them. This will ensure that non-Microsoft middleware is able to interoperate with the operating system and compete on a function-by-function basis with Microsoft middleware. Microsoft must also allow computer manufacturers and end users to replace Microsoft middleware and preserve "default" settings that will ensure that Microsoft's middleware does not override the selection of competing middleware products. The consent decree also requires Microsoft to license communications protocols that allow its servers to interoperate with the operating system, given the potential threat that server-based applications may pose to the Windows monopoly.

On the issue of enforcement, some have suggested that no conduct remedy would be effective in this case because Microsoft is defiant and cannot be trusted. Our practice in enforcement, however, is never influenced by the degree to which we "trust" a defendant. We assume that companies will always pursue their own best interests, which are not necessarily coincident with the public interest, and we fashion remedies accordingly. Our decree includes some of the most stringent enforcement provisions ever included in a decree. In addition to ordinary prosecutorial access powers, backed by criminal and civil contempt authority, the decree has two aggressive features: (1) an independent, full-time, on-site compliance team, including staff and consultants, with the authority to monitor compliance, report violations, attempt to resolve technical disputes, and gain full access to Microsoft's business, employees, and records, including its source code; and (2) extension of the decree's term by up to two years if the court finds willful, systematic violations of the decree.

Pursuant to agreement with the Department, Microsoft began complying with the terms of the decree late last year, even before its entry (a common practice when we enter into consent decrees). We understand that participants in the software industry have already begun to take advantage of the terms of the decree.

There are those who complain that we did not do enough to "punish" Microsoft. Others complain about the lack of any monetary sanction. These should not be debated seriously in antitrust circles, because, of course, we have no authority to punish a civil defendant or to seek fines or damages. Others complain that we should have taken away Microsoft's monopoly. But the fact is that we never alleged in our case that Microsoft had unlawfully obtained its monopoly. And, of course, we had the Court of Appeals admonishing us not to seek a structural remedy without establishing a more direct causal connection between the unlawful acts and the maintenance of the monopoly.

Another common complaint (so common that when it has been made and we have asked what complainants are referring to, they often cannot point to anything specific), which I mentioned above, is that the decree is laden with so-called "loopholes." It is important to focus closely on those "loopholes." I believe the critics are referring to places in the decree in which we make clear that certain prohibitions on discrimination are not intended to prevent Microsoft from engaging in lawful collaborative conduct. The fact is that using a wholly blunt instrument to place prohibitions on Microsoft would prevent them also from engaging in lawful, potentially procompetitive behavior, prevention of which could adversely affect many of its business partners and the market itself. We believe we have an obligation as enforcers to take a look at the unintended consequences that broad prohibitions would have and try to direct them as specifically at the unlawful conduct as possible. Had we not allowed such carve-outs for lawful behavior, we would have had a line at the door from companies wanting to partner with Microsoft saying: "surely you did not mean to prevent Microsoft from this kind of behavior! Come on, give them assurance that if they collaborate with us and give us the corresponding access to their products or knowledge that you will not seek a contempt finding even though the technical language of the decree would require it." How can we be so sure that we would have received such requests? Not only were we fairly certain that would be the case when we agreed to the provisions, but, in fact, it now has happened. Companies that have criticized us for loopholes have already called to seek some!

The fact is that blunt instruments in remedying antitrust violations can be dangerous. In the cross-examination of Microsoft Chairman Bill Gates in the states' remedies trial, the states' attorneys seemed to try to get him to admit that even though language in some of the states' proposed remedy provisions might technically be quite broad, no state would ever interpret it so broadly in enforcement. That would be poor remedy practice. It is hard to imagine that responsible decree enforcement should ever begin with the implication -- if not bare assertion -- that even though the language is broad, a defendant should feel free to take action prohibited by the language and trust that the enforcer will simply be reasonable and sort out later what is really prohibited.

The loudest critics are Microsoft's competitors. Many of them provided valuable assistance to us during the liability phase. And their opinions were made known to us during the remedy phase. But we had a sense that, by that point, a sense of entitlement had set in. Requests were made that we agree to remedies that, in our view, would have hindered or prevented Microsoft from competing, while greatly advantaging its competitors -- including in products and markets with only the most indirect relationship to maintenance of the operating system monopoly. We received and reviewed a host of "wish lists." Some of it was helpful; some of it was pure rent-seeking. (And, indeed, in the non-settling states' remedy trial, evidence was presented that some companies sought advantageous deals with Microsoft in exchange for telling the Department positive things about Microsoft's behavior.)

At the end of their testimony for the states, Microsoft's competitors often repeated a common refrain, plainly in response to Microsoft's impeachment attempts, fairly paraphrased as: "We aren't trying to secure a tactical advantage; we just want a level playing field." Now that may have some gut-level appeal in a courtroom, particularly if there were a jury present (which there was not). And I have no doubt that these witnesses and their companies are sincere in wanting a "level playing field." But that is not what the U.S. antitrust laws provide. Outside of a regulatory context, there is no mandate under Section 2 of the Sherman Act to "level the playing field." You will find the mandate to prevent someone from using exclusionary conduct to put a fence around the field. You will find a mandate to prevent someone from predatory conduct in bulldozing everyone else off the field. And you will find a mandate to remedy the unlawful conduct when it is found. But the surface of the playing field is to be determined through competition. This is especially the case here in Microsoft because we and the states did not allege that Microsoft unlawfully obtained its monopoly position.

We never doubt for a moment that third-party companies will behave strategically; their legal obligations, after all, are to shareholders, not to the public at large. But our obligation is to the public at large. And this serves as an important reminder to us that we must make our own judgments, based upon our own assessment of the market place, and that our job is not to regulate competition in the software industry.

The fact that the Department of Justice and some states diverged in their views on remedying Microsoft's violations has fueled the debate over the effectiveness and efficiency of our multi-faceted antitrust enforcement scheme. The action by the nine non-settling states raises for the first time the prospect that a small group of states, with no particularized interests to vindicate, could obtain divergent relief with wide-ranging, national and even international economic implications. In the states' case, the district court requested that the Division weigh in on the issue of whether the non-settling states had standing to pursue such relief, given that the United States had taken a different position -- an issue that Microsoft had raised in a motion to dismiss.⁽¹⁶⁾ In our amicus brief, we made several points.⁽¹⁷⁾ First, the United States is the sole enforcer of the federal antitrust laws on behalf of the American public. While the states have the authority to seek injunctive relief under federal law, they do so as private parties under Section 16 of the Clayton Act, 15 U.S.C. § 26, not as sovereign law enforcers. Second, nonetheless, precedent does not require dismissal of the states' case as a matter of law.

Finally, and perhaps most significantly, we stated that the limitations of Clayton Act Section 16 and the enforcement judgement of the United States are important considerations in the Court's exercise of equitable discretion. We took the position, backed by case law, that the Court in exercising equitable discretion should take into account the United States' enforcement judgment in entering into its settlement and, if approved, how that bears on the non-settling states' requisite threatened loss or damage; whether a small group of states are the parties best situated to obtain relief of such broad reach and implication; and whether the states are using their parens patriae authority improperly to act on behalf of the private interests of certain firms.

On June 12, 2002, the Court denied Microsoft's motion to dismiss, largely on the basis that the Court of Appeals had implicitly or explicitly confirmed the non-settling states' parens patriae standing.⁽¹⁸⁾ The Court did state, however, that the policy arguments in the United States' amicus brief may inform the Court's exercise of its equitable powers in devising a remedy in the case.⁽¹⁹⁾

FOOTNOTES

1. Transcript of Roundtable Conference with Enforcement Officials, 70 Antitrust L.J. 261, 297 (2002).

2. Prepared Remarks of Timothy J. Muris, "Merger Enforcement in a World of Multiple Arbiters," December 21, 2001; see http://www.ftc.gov/speeches/muris/brookings.pdf.

3. Civil Action No. 98-1232 (D.D.C. filed May 18, 1998).

4. See http://www.ftc.gov/os/1999/9908/divestiture.pdf.

5. United States v. Microsoft Corp., 253 F.3d 34 (D.C. Cir. 2001).

6. State of New York, et al. v. Microsoft Corp., Civil Action No. 98-1233 (D.D.C. filed May 18, 1998).

7. United States v. Microsoft Corp., 1998-2 Trade Cases ¶ 72,261 (D.D.C. 1998).

8. United States v. Microsoft Corp., 87 F. Supp.2d 30 (D.D.C. 2000).

9. United States v. Microsoft Corp., 97 F. Supp.2d 59 (D.D.C. 2000).

10. Microsoft, 253 F.3d 34.

11. Id. at 105.

12. Order, United States v. Microsoft Corp., at 2-3 (September 28, 2001).

13. See Revised Proposed Final Judgment, United States v. Microsoft Corp. (filed November 6, 2001) (http://www.usdoj.gov/atr/cases/f9400/9495.htm) and Second Revised Proposed Final Judgment (filed on February 27, 2002) (http://www.usdoj.gov/atr/cases/f10100/10146a.htm).

14. Response of the United States to Public Comments on the Revised Proposed Final Judgment, United States v. Microsoft Corp. (filed February 27, 2002) (http://www.usdoj.gov/atr/ cases f10100/10145.htm).

15. United States v. Microsoft, 2002 WL 1676762 (D.D.C. 2002).

16. Motion for Dismissal of the Non-Settling States' Demand for Equitable Relief, State of New York, et al. v. Microsoft Corp. (filed February 26, 2002); Order, State of New York, et al. v. Microsoft Corp. (March 25, 2002).

17. Memorandum Amicus Curiae of the United States of America Regarding Microsoft Corporation's Motion for Dismissal of the Non-Settling States' Demand for Equitable Relief, State of New York, et al. v. Microsoft Corp. (filed April 15, 2002); see http://www.usdoj.gov/atr/ cases/f10900/10980.htm.

18. State of New York, et al. v. Microsoft Corp., 2002 WL 1311437 (D.D.C. 2002).

19. Id. at *97 n.28.