Microsoft Tunney Act Comment : The Honorable Herb Kohl, U.S. Senator

VIA FACSIMILE (202) 307-1454

The Honorable Charles James
Assistant Attorney General for Antitrust
c/o Renata Hesse, Antitrust Division
U.S. Department of Justice
601 D Street, N-W, Suite 1200
Washington, DC 20530-0001

Dear Assistant Attorney General James:

The proposed Final Judgment submitted in U.S. v. Microsoft Corporation, and State of New York et al. v. Microsoft Corporation, triggers obligations by you under the Antitrust Procedures and Penalties Act (the "Tunney Act"), set forth at 15 U.S.C. § 16. Under the Act, you are designated to "receive and consider any written comments relating to the proposal for [a] consent judgment submitted under... this section." i5 U.S.C. §16(d). On November 8, 2001, Judge Colleen Kollar-Kotelly ordered "that members of the public may submit written comments concerning the proposed Final Judgment to [the Justice Department.]" This letter is sent to you pursuant to this statute and court order.

As Chairman of the Subcommittee on Antitrust, Business Rights, and Competition of the Senate Committee on the Judiciary, I have studied the proposed settlement of the government's antitrust lawsuit against Microsoft very closely, and I write to express my concern about whether the settlement is in fact "in the public interest." 15 U.S.C. § 16(e). Accordingly, I respectfully ask that you address the issues raised in this letter when you file with the district court your mandatory "response" to these comments. See 15 U.S.C. § 16(d).

This settlement affects vast and important segments of the U.S. economy, and thus, its significance cannot be overstated. As a result, I believe it should only be approved if it can truly be shown that the settlement is "is m the public interest," Such a determination will require the court to assess "the competitive impact of such judgment, including termination of alleged violations, provisions for enforcement and modification, duration of relief sought, [and] anticipated effects of alternative remedies actually considered." 15 U.S.C. § 16(e)(1). These comments and questions are designed to inform this inquiry.

(1) Does the proposed settlement contain significant loopholes that render it largely ineffective to cure the damage to competition caused by Microsoft's illegal behavior? The unanimous District of Columbia Circuit Court of Appeals held that Microsoft violated section 2 of the Sherman Act by illegal conduct designed to maintain its monopoly on personal computer operating systems. The proposed settlement is designed to "provide a prompt, certain and effective remedy for consumers" arid "halt continuance mad prevent recurrence of the violations of the Sherman Act by Microsoft... and restore competitive condition, s to the market." Competitive Impact Statement at 2.

However, it appears that, in many respects, the Revised Proposed Final Judgment ("PFJ"), contains so many loopholes, exceptions, qualifications, and definitional limitations that Microsoft can easily avoid its requirements. I have serious concerns that these loopholes and qualifications in the proposed settlement render it inadequate to accomplish its task of remedying Microsoft's illegal conduct and restoring competition in the computer software market. I will list a few examples below, but this list is not exhaustive.

(a) Million software copy limitation - Under the proposed consent decree, competitive access to the computer desktop has to be provided for certain types of firs.1 software application makers. More specifically, Microsoft must permit computer manufacturers and computer users to replace the icons, short-cuts, or menu entries for Microsoft Middleware Products on the desktop or start menu with Non-Microsoft Middleware Products. See PFJ ¶ III.H.1. But the proposed consent decree contains a loophole in the definition of Non-Microsoft Middleware Products that denies that protection unless, in the prior year, "at least one million copies [of that rival software] were distributed m the United States." Id. ¶ VI.N. Thus, many start-up software companies with promising yet unproven technology in the pipe]me - precisely the companies most in need of protection from exercise of market power by a monopolist - will be left unprotected by the settlement. This could have a negative impact on the flow of venture capital and investment to technology start-ups - precisely the engine that drove the economic expansion of the late 1990s and a key to further expansion of our all-important technology sector.

Requiring the distribution of a million copies in the United States in a year seems a very high threshhold. Why was this limitation written into the settlement? Why was the one million number chosen? In this era of internet downloads, how can a software maker prove that a million copies were distributed in the United States? What purpose does this limitation serve? Should the consent decree be modified to close that loophole and foster new investments and growth by protecting those investments from monopoly practices?

In answering this inquiry I would request that you conduct a survey of a representative sample of non-Microsoft middleware product manufacturers to determine how long it took them to distribute one million copies of their software in the United States (if in fact they have done so). For example, it would be fruitful to determine how long it took products such as Kodak photofinishing software or the Palm OS to reach this threshhold, even though they might not qualify as non-Microsoft middleware products under the decree.

(b) Definitional limits on API disclosures - A cornerstone of the proposed settlement is the requirement that Microsoft disclose certain Application Programming Interfaces ("APIs") that are used by Microsoft Middleware to interoperate with a Windows Operating System Product. See ¶ III.D. Yet the definition of Microsoft Middleware is limited in such a way that raises doubts about the true extent of the requirement of API disclosure. Microsoft Middleware must be distributed seperately from a Windows Operating System Product and must be trademarked. Id. ¶ VI.J. And any product comprised of the Microsoft or Windows trademark together with descriptive or generic terms are not considered to be trademarked. Id. ¶ VI.T.

It appears that these definitional limitations will make it easy for Microsoft to avoid having to disclose APIs. Why can Microsoft simply decide not to trademark software and thereby have it fall outside this definition? Why does the definition of trademark exclude products identified by the Microsoft or Windows trademark plus a generic term?

In addition, Microsoft need not release any API to a software maker unless Microsoft determines that the software maker "has a reasonable business need for the API" and "meets reasonable, objective standards established by Microsoft for certifying the authenticity and viability of its business." Id. ¶ III.J. Many are concerned that permitting Microsoft to determine the "business need" and "viability" of potential competitors will be another way Microsoft can avoid the API disclosure requirements.
Another limit on API disclosure applies to new versions of Windows Operating Systems Products. With respect to new versions, disclosures of APIs need only be made in a Timely Manner. Id. ¶ III.D. Timely Manner is defined to mean when Microsoft releases a test version of a Windows Operating System Product to 150,000 beta testers. Id. § VI.R. Microsoft could avoid this provision by simply having the new version - and the obligation to release APIs tested by less than 150,000 beta testers. What assurance is there that Microsoft will not avoid the API disclosure limitation in the future in this manner?

There are many other definitional limits and qualifications found throughout the proposed consent decree. Beyond their specific provisions, these limitations raise the broader - and fundamental - question: is the proposed settlement is strong enough to make sure that Microsoft cannot use its monopoly power to squelch competition and innovation?

(2) Is the scope of the settlement's protection of middleware adequate to promote competition ? Before even running the gauntlet of the decree's restrictions outlined above, software must qualify as non-Microsoft middleware under the restrictive definitions of middleware used in the decree, see PFJ ¶¶ VI.J, VI.K, VI.M, VI.N. The decree's protections are largely limited to competitors of "Microsoft Middleware Products." But the definition of what is (or is not) a Microsoft Middleware Product remains somewhat ambiguous. Windows Messenger is covered, but MSN Messenger does not appear to be. Internet Explorer is covered, but MSN Explorer seems to be missing. Popular products such as software for personal digital assistants and photofinishing are excluded. What accounts for these gaps? And why did the Department of Justice and Microsoft abandon the definition of middleware that had been employed by, the District Court and affirmed unanimously on appeal?

These definitions lie at the core of the consent decree's potential effectiveness. The heart of the Court of Appeals' ruling was that Microsoft's acts of illegal monopoly maintenance blocked the ability of competitors to develop middleware which could effectively become an alternative platform to compete with the Windows operating system. The goal of the consent decree therefore must be to encourage and protect innovation in the middleware field. Yet the more restrictive and unclear the definitions are, the more they introduce uncertainty into this field and the more ineffective the consent decree will be. If potential innovators believe that Microsoft can avoid the ambit of the decree, then hopes for spurring innovation and competition among middleware products will be lost.

(3) Why is Microsoft ever allowed to retaliate against the computer makers? The settlement rests on the computer makers' ability to promote competition on the desktop and in the industry generally. A key provision of the consent decree bans Microsoft from retaliating by agreement for a computer makers loading certain types of non-Microsoft software on its computers. See PFJ ¶ III.A. Yet, the ban on Microsoft retaliation against computer makers is limited: the decree only bans retaliation in commercial agreements (not other forms of retaliation); only bans retaliation for removing specifically named "Microsoft Middleware Products" (not other Microsoft products); and only bans retaliation, for promoting specific competitive products (and not other products that could challenge Microsoft's desktop dominance). Will these loopholes "swallow the rule" that Microsoft should be banned from retaliating against computer makers? Would it not be far simpler to ban all Microsoft retaliation against the computer makers?

(4) Will Microsoft be able to accomplish through incentives what it could not accomplish by retaliation? The ban on retaliation in no way prevents Microsoft from paying incentives to computer makers to strongly prefer - or install exclusively - Microsoft software products. Indeed, the proposed consent decree expressly provides that nothing "shall prohibit Microsoft from providing Consideration to any OEM with respect to any Microsoft product or service where that Consideration is commensurate with the absolute level of that OEM's development, distribution, promotion, or licensing of that Microsoft product or service." PFJ ¶ III.A. Given Microsoft's market power and financial resources, what will prevent Microsoft from using its financial resources, what will prevent Microsoft from paying bounties to computer manufacturers to "voluntarily" exclusively install or at least to prefer, Microsoft products, thereby accomplishing through incentive payments what it could not achieve by retaliation?

(5) Why does the settlement abandon the ban on commingling that the Court of Appeals found to be illegal? Nothing in the agreement prohibits Microsoft from commingling code or binding of its middleware to the Operating System (OS), even though the Court of Appeals specifically found Microsoft's commingling of browser and OS code to be anti-competitive and rejected a Microsoft petition for rehearing that centered on this issue. Computer manufacturers are likely to be discouraged from installing competing middleware products to those commingled with OS code, as these are likely to slow down the computer's speed and performance. Why should the proposed settlement permit this commingling to continue in the face of an explicit finding of illegality from the Court of Appeals?

(6) Is the five year term of the settlement sufficient to restore competition? The proposed consent decree has a five year term (extendable for two years only if the Court finds Microsoft has engaged in willful and systematic violations of the decree). PFJ ¶ V. This is an unusually short time for an antitrust consent decree, which is typically ten years in length. Many wonder if this term is sufficient to remedy Microsoft's illegal conduct and restore competition. Why was the term of the decree limited to five years? How can we be sure that five years will be sufficient to restore competition?

In addition, under the decree, Microsoft has up to a year after submission of the decree before implementing several of its provisions, including the crucial API disclosure requirements and the provisions granting computer manufacturers and users the right to modify the start-up menus and icons with competing products. See PFJ ¶¶ III.D, III.H. Thus the effective time that Microsoft must live under these restrictions is substantially shorter than the five year term of the agreement. And yet, this summer, when Microsoft made some very, minor changes to Windows to respond to the Court of Appeals ruling, it took just three weeks to make the changes. Given Microsoft's proven ability to make rapid changes, would it not be in the public interest to require Microsoft to live by the consent decree immediately and not wait another year?

(7) Is the enforcement mechanism sufficient? The proposed Final Judgment does not set forth vigorous enforcement mechanisms to keep Microsoft within the framework of this settlement. The proposed consent decree requirements the appointment of a "Technical Committee" to monitor compliance with the decree, but its findings are entirely advisory and not binding on Microsoft. PFJ ¶ IV.B. The only true enforcement mechanism would be for the Justice Department to go to Court in art enforcement action. In such an enforcement action, no work product findings or recommendations of the Technical Committee can be admitted as evidence in court. Id. ¶ IV.D.4.d. "While Microsoft is required to "be reasonable" in its conduct: violations of such "be reasonable" provisions can only be remedied through proceedings that will become, in essence, mini-retrials of U.S. v. Microsoft itself. Are these provisions sufficient to ensure that the settlement can be enforced properly? Without an iron-clad enforcement mechanism, how can the public take solace in the "promise" that Microsoft will "be reasonable" given the history of litigation in this case, and earlier antitrust lawsuits against Microsoft?

(8) What will the settlement's effect be on Microsoft's future conduct? Microsoft is has dubbed its aggressive Internet strategy .NET or "Hailstorm" - a strategy to give consumers a one-stop shop on the Internet. How will the consent decree foster competition for these future "platforms?" If the purpose of this case was to check Microsoft's monopoly power, how will the Justice Department ensure that this monopoly dominance is not extended from the desktop to the Internet? And why are critical new technologies, such as digital rights management and identity-authentication, exempted from the proposed settlement's disclosure provisions?

In closing, we today stand on the threshold of writing the rules for competition in the digital age, and we have two choices. One option involves one dominant company controlling the computer desktop, facing minor restraints that expire in five years, but acting as a gatekeeper to 95% of all personal computer users. The other model is the flowering of innovation and new products that resulted from the ending of the AT&T telephone monopoly nearly twenty, years ago. From cell phones to faxes, from long distance price wars to the development of the Interact itself, the end of the telephone monopoly brought an explosion of new technologies and services that benefit millions of consumers every day. We should resist on nothing less from this proposed settlement.

In sum, any settlement in this case should make the market for computer software at least as competitive as the market for computer hardware is today. We should insist on a settlement that has an immediate, substantial, and permanent impact on restoring competition in this industry.

I recognize the extraordinary effort that the Justice Department has expended in the litigation of this case, and I thank you and your staff for the vigor with which you have pursued this challenging case. I believe that answers to the questions and issues posed above are essential for determining whether the proposed settlement is in the public interest.
Thank you for your attention to this matter.