IN
THE UNITED STATES DISTRICT COURT
FOR
THE DISTRICT OF COLUMBIA
________________________________________________
UNITED STATES OF AMERICA, |
Plaintiff,
Defendant. |
__________________________________________
Attorney General ELIOT SPITZER, et al.,
Plaintiffs,
Defendant.
COMPUTER
AND COMMUNICATIONS INDUSTRY ASSOCIATION AND
SOFTWARE
AND INFORMATION INDUSTRY ASSOCIATION
Edward J. Black, President (Bar No. 113282)
Ken
Wasch, President (Bar No. 934984)
Jason Mahler (Bar No. 435605)
Software and Information
Computer & Communications
Industry Association
Industry Association
1730 M Street, N.W
666 11th Street, N.W.
Washington, D.C. 20036
Washington, D.C. 20001
(202) 451-1600
(202) 783-0070
Page
They
Reflect A Serious Structural Problem...................................................
6
Threat
Or To Reinvigorate Competition..............................................................................
11
TABLE
OF AUTHORITIES
Page(s)
Cases
(7th Cir. 1998), cert. denied, 525 U.S. 1071
(1999) ...........................................................
7
(D.C. Cir. 1995) ........................................................................................................
22, 37
(D.D.C.) ..........................................................................................................................
13
464 U.S. 1013 (1983) ......................................................................................................
47
Remarks of
Charles F. Rule before The Brookings Institution,
Developments in Telecommunications Policy, Oct. 5,
1988.................................................
20
(rev. ed. 1996 & 1999 supp.) ..................................................................................... passim
http://www.zdnet.com/pcweek/stories/news/0,4153,1018247,00.html.......................... 19, 46
http://www.microsoft.com/msft/speech/analystmtg99/ballmerfam99.html.......................... 7, 13
Apr. 25, 2000,
http://www.microsoft.com/billgates/speeches/04?25winhec00.htm................ 7
http://www.zdnet.com/sr/stories/issue/0,4537,392493,00.html............................................
16
May 1, 2000, at A34...................................................................................................
33, 45
(D.C. Cir. filed July 25, 1989)...........................................................................................
52
No. 725, Oct. Term, 1909...........................................................................................
41-42
Remedy Sought, Seattle Post-Intelligencer, Nov. 10,
1999, at A1.................
48
at C4.................................................................................................................................
50
Times-Picayune, May 2, 2000, at C1........................................................................
48
Aims to Neutralize Antitrust Efforts, Wash.
Post, May 7, 1999, at A1 .....................
21
Act Monopolization Cases, 2 Int’l
J. Econ. & Bus. 263 (1995) ...............................
52
Rating and $145 Price Target, Jan. 14, 2000..................................................................
53
http://www.zdnet.com/pcmag/stories/opinions/0,7802,2559857,00.html............................. 45
June 2, 1999, http://www.zdnet.com/sr/stories/0,4538,
2268932,00.html...........................
16
http://www.zdnet.com/zdnn/stories/news/0,4586,2556472,00.html.....................................
16
Growth
in First Quarter 2000, press release,
Apr. 24, 2000,
http://gartner11.gartnerweb.com/dq/static/about/press/pr?b200019.html................................ 7
Microsoft
Plows Ahead with its Plan to Dominate the Internet, Industry
Standard, May 22, 2000, at 120................................................................................
19
http://www.mercurycenter.com/svtech/columns/gillmor/docs/dg042300.htm........................ 46
(San Jose
Mercury News) Apr. 24, 2000,
http://weblog.mercurycenter.com/ejournal/2000/04/24........................................................
43
Resellers
Broaden Scope of Enterprise Agreements, Computer Reseller
News, Apr. 20, 1998,
http://www.techweb.com/se/directlink.cgi?
CRN19980420S0003.......................................................................................................
14
Nov. 5, 1995, at 50...........................................................................................................
38
at 48...........................................................................................................................
37, 40
Apr. 26, 2000, at E1 ........................................................................................................
38
April 24, 2000, at A1 .......................................................................................................
33
to Change, Wash. Post, Apr. 19, 2000, at E1 ......................................................
17, 23
Seattle Times, Feb. 17, 1999, at A1 .........................................................................
21
May 17, 2000, at A1.........................................................................................................
22
31 Conn.
L. Rev. 1285 (1999)...............................................................................
41, 51
at A26.................................................................................................................................
7
Will Be No Breakup, N.Y. Times, Apr.
26, 2000, at C1...........................................
23, 43
Apr. 5, 2000.....................................................................................................................
52
74 Antitrust
& Trade Reg. Rep. (BNA) 49 (Jan. 15, 1998).................................. 21
licensing/agree.htm (visited May 2000).........................................................................
13, 14
1998,
http://www.microsoft.com/enterprise/licensing/agreement/Eawhite.doc................ 13, 14
Inc., Nov. 11, 1999, http://www.g2news.com....................................................................
45
To Quit Lawsuit Coincidental, Chi.
Trib., Dec. 25, 1998, at 3 ....................................
21
http://www.techweb.com/wire/story/TWB19991203S0020......................................... passim
zdnn/stories/news/0,4586,2560616,00.html ......................................................................
23
Pieces Would Be Enough?, S.J.
Mercury News, May 1, 2000.................................
46
Outcomes, International
Data Corp. (1999)........................................................................ 52
No. 398, Oct. Term, 1910..........................................................................................
42, 51
Dallas Morning News, May 4, 2000, at 3F..........................................................
46
Observers, InfoWorld.com, Apr. 28, 2000,
http://www.infoworld.com/
articles/pi/xml/00/04/28/000428pivendorreax.xml...............................................................
49
at E1.................................................................................................................................
49
News Service, March 11, 1999.........................................................................................
21
May 15, 2000, at B1.........................................................................................................
22
Antitrust Case, Wall St. J., May 16, 2000, at A3........................................................
22
INTEREST OF THE AMICI CURIAE
The Computer & Communications Industry Association
(CCIA) and the Software and Information Industry Association (SIIA) are
technology trade associations. Each has a long pedigree, broad membership and a
record of participation in antitrust matters involving Microsoft. CCIA was one
of the principal amici in the review of the current consent decree in both this
Court and the court of appeals. SIIA has appeared in this proceeding as amicus
curiae supporting the United States at the liability stage. In this brief, the
two organizations present their views on an appropriate remedy.
1. The
Computer & Communications Industry Association is an association of
computer technology and telecommunications companies that range from small
entrepreneurial firms to some of the largest members of the industry. CCIA’s
members include equipment manufacturers, software developers, providers of
electronic commerce, networking, telecommunications and on-line services,
resellers, systems integrators, and third-party vendors. Its member companies
employ nearly one million persons and generate annual revenues exceeding $300
billion. CCIA’s mission is to further the business interests of its members,
their customers, and the industry at large by being the leading industry
advocate in promoting open, barrier-free competition in the offering of
computer and communications products and services worldwide. CCIA’s motto is
“Open Markets, Open Systems, Open Networks, and Full, Fair and Open
Competition,” and its website is at www.ccianet.org.
For more than 26 years, CCIA has supported antitrust
policy that ensures competition and a level playing field in the computer industry.
CCIA supported the Tunney Act in the 1973 congressional hearings preceding the
enactment of that legislation, and participated as amicus curiae in the
proceedings examining the current Microsoft consent decree. CCIA is intimately
familiar with the shortcomings of that decree, and its failure to prevent or
deter Microsoft from continuing on an anticompetitive course.
2. The
Software and Information Industry Association is the world’s largest trade
association representing the interests of firms in the software, information
and Internet industry. Formed on January 1, 1999, through the merger of the
15-year-old Software Publishers Association (SPA) and the 30-year-old
Information Industry Association, SIIA leads industry efforts in e-business,
copyright, privacy, taxation and other public policy issues. SIIA’s website is
at www.siia.net.
3. CCIA,
SIIA, and their members participate in all aspects of the computer software,
information, communications, and Internet industries. They have a vital interest
in the outcome of this proceeding because the future structure of the computer
software industry and of Internet computing — and the range of conduct that the
law permits within it — will determine to a substantial extent whether they
thrive in a fair, innovative, and competitive environment. Competitive markets
produce the greatest amount of innovation and provide consumers with the best
products at the lowest prices. Just as those results benefit consumers, they
are in the best interests of the industry.
CCIA, SIIA, and their members are thoroughly familiar
with the markets and practices at issue in this case, and with the practical
significance of that conduct. SIIA and CCIA reject the notion that the software
industry can function efficiently only under rules prescribed by a monopolist
who dictates when innovation may take place, what form it may take, and who may
engage in it. Like the computer, communications and content industries, the
software industry functions best when companies within it are free to engage in
a dynamic and unrestrained competitive process. In competitive markets,
innovation occurs and software interoperates smoothly without Microsoft’s
governance.
4. Although
Microsoft used to be a member of SIIA — and a member of the SIIA Board of
Directors — Microsoft resigned from SIIA and withdrew its funding after SIIA
filed an amicus brief criticizing Microsoft’s conduct at the liability stage of
this proceeding. Microsoft has also induced some other companies dependent upon
it to withdraw funding from both amici. These events shed a strong light on the
remedy issue now before the Court. Microsoft’s power and wealth give it the
ability to both punish its critics and retain battalions of lawyers, lobbyists,
and publicists to undermine the government at every turn. If the Court adopts a
“good conduct” remedy in this case, it can be sure that Microsoft will seek to
silence those who would inform the decree court of future infractions, and will
dispute the meaning of every provision in the decree — just as it did after the
entry of the consent decree in 1995 — with the predictable consequence that the
decree will be reduced to a dead letter. As we demonstrate in this brief,
Microsoft is too powerful to be “fenced in” with a good conduct code, no matter
how carefully that code is written.
INTRODUCTION
AND SUMMARY OF ARGUMENT
There is little dispute that this is the most
important antitrust case of our generation, the case that will determine whether
the structure of software markets — and the progress of the 21st Century
economy — will be based on competition or monopoly. Few, if any, issues will
affect consumers more in the coming years. Because the formulation of a remedy
for antitrust violations is the “most significant phase of the case,” United
States v. Glaxo Group Ltd., 410 U.S. 52, 64 (1973), it is crucial
that the Court arrive at its judgment with both care and speed. Software
markets move quickly; if permitted to continue on its course unimpeded,
Microsoft can consolidate the rewards of its anticompetitive conduct quickly as
well. The plaintiffs’ evidence shows that, after more than a decade of
antitrust enforcement scrutiny, Microsoft continues to use illegal means to
short circuit competitive challenges to its dominance. It is time for those
abuses to end.
A. As
this Court has recognized, the structural problem that has given force to
Microsoft’s anticompetitive abuses is the applications barrier to entry. That
is the structural condition that Microsoft has exploited (and has reinforced
and preserved). No list of behavioral proscriptions could effectively contain
that exploitation. Microsoft has shown remarkable inventiveness in devising new
ways to leverage its market power to foreclose competition. Accordingly, any
effective remedy must attack this problem, whether directly (as proposed in the
Remedies Brief of Amici Curiae Robert Litan et al. (“Economists’
Brief”)), or indirectly, as the plaintiffs propose.
A structural remedy is needed because behavioral
remedies do not address the two principal competitive problems demonstrated by
the trial evidence and subsequent events:(1) Microsoft’s monopoly power in
operating systems, which provides multifaceted opportunities for abusive, coercive
conduct that excludes competition, and (2) Microsoft’s successful leveraging of
that monopoly into a monopoly in the Internet browser. The latter monopoly
provides a chokehold over Internet computing that permits Microsoft to
transform open-standard Internet computing into a Microsoft-proprietary domain.
As the Court is well aware, if a remedy requires future, reactive enforcement
proceedings to impose tangible constraints on Microsoft, Microsoft will use
those proceedings as an opportunity for gamesmanship and obstruction — annexing
additional markets to its monopoly in the meantime, as it has done while the
1995 consent decree has been in force. Plaintiffs’ Proposed Final Judgment
provides a measured but effective way of dealing with these problems, and does
so without the delay and administrative intrusion occasioned by a remedy that
relies on conduct restraints alone.
Plaintiffs’ Proposed Final Judgment accords with the
remedial principles laid down by the Supreme Court in antitrust case after antitrust
case. Antitrust remedies must restore competition, neutralize a monopoly that
has been abused, deprive a violator of the benefits of its illegal conduct, and
prevent a recurrence of anticompetitive activity. Where the antitrust violation
involves monopoly, and there is a continuing incentive and ability to abuse
that monopoly, only a structural remedy can satisfy these criteria.
B. Plaintiffs’
Proposed Final Judgment provides an appropriate framework for relief on the
facts of this case. The reorganization of Microsoft into separate operating
systems and applications businesses takes away the operating system’s control
over the most significant aspects of the applications barrier to entry —
including control over the Internet browser achieved by Microsoft’s illegal
conduct proved at trial. Although the operating systems company will remain a
monopoly immediately after the reorganization, it will face the Office and
Internet Explorer monopolies as competitive threats rather than reinforcements
of its power. Each of the companies to be formed in the reorganization will
have incentives to undermine the other’s monopoly control, whether by entering
the other’s market, by ensuring interoperability with the other’s rivals, or
merely by cooperating in the development of cross-platform middleware aimed at
the other’s platform. Consumers benefit from the unleashing of both actual and
potential competition.
The conduct restrictions in Plaintiffs’ Proposed Final
Judgment do not prevent either successor company from engaging in any line of
business, and thus avoid the principal drawback of the AT&T decree.
The proposed conduct provisions instead are appropriate interim measures to
prevent Microsoft from engaging in anticompetitive behavior similar to that
proved in this case, without impinging on the development of software. Given
Microsoft’s track record, such restraints are needed until the reorganized,
independent successor companies establish a competitive dynamic.
This Court need not delay final resolution of this
case if it enters Plaintiffs’ Proposed Final Judgment. That Judgment would be
final for the purposes of an expedited appeal to the Supreme Court, and the
details of implementation can be formulated while the appeal is pending. Before
judgment is entered, Microsoft may be afforded an opportunity to cross-examine
the plaintiffs’ witnesses promptly, but should not receive additional
discovery. Microsoft knows its own organizational structure and its own illegal
practices. It should not be permitted to use the judicial process to seek out
and intimidate those who provided information to the government, in hopes of
deterring assistance by customers and others in the enforcement of the decree
and in the reporting of future violations.
C. Although
Plaintiffs’ Proposed Final Judgment goes far toward undoing the competitive
harm caused by Microsoft’s widespread monopolistic abuses, a small but
competitively significant adjustment would make the remedy more robust. The
Court should supplement the proposed reorganization with a provision separating
the Internet Explorer intellectual property and associated personnel into a
separate company (with a license of the current Internet Explorer product to
the operating systems and applications companies). In the alternative, the
Court should order that the applications company make the Internet Explorer
product — which provides no royalties now — an “open source” product so that
other software developers could use the source code. Either of these small additions
would ensure that the monopoly over productivity applications that Microsoft
holds does not supplant the operating system as the point of leverage for a
monopoly over the software used in Internet computing.
D. At
the remedy stage of this case, the danger is not of doing too much but of doing
too little. This is the most significant monopolization violation proved in a
generation, and the most significant monopolization case litigated to judgment
in many decades, in an industry of surpassing importance to the current and
future national economy. If Microsoft walks away from this case with
only another set of conduct restrictions to evade, Section 2 of the Sherman Act
will be drained of much of its practical effect. That is a price the public
should not be asked to pay.
ARGUMENT
MICROSOFT SHOULD
BE REORGANIZED TO REDUCE ITS MONOPOLY POWER
A. Microsoft’s
Violations Call For Structural Relief Because They Reflect A Serious Structural
Problem
1. Microsoft’s
Violations Reflect Pervasive and Multifaceted Abuses of Monopoly Power
a. This
case involves the monopolization of the software central to current desktop
computing — the operating system — and the attempted (now, successful) monopolization
of the software that is central to Internet computing — the Internet browser.
Microsoft’s anticompetitive conduct was not narrow and confined. Rather,
Microsoft aimed its pervasive illegal practices at any product or firm that
presented even a potential threat to Microsoft’s monopoly power. That conduct
was not the work of a few rogue employees without corporate authority or
control. To the contrary, anticompetitive conduct was orchestrated at the
highest levels of the company — often under the direction of Bill Gates
himself, a state of affairs that has continued since the evidence closed. See
Henderson Dec. ¶ 69 (citing Gov’t Remedy Ex. 1).
Moreover, despite Microsoft’s recurrent contentions,
its monopolistic abuses harmed consumers. See Findings ¶¶ 57, 60, 62-66,
171-174, 210-216, 225-229, 247, 339-340, 379, 397, 408-412.Consumers have lost
the price and functionality benefits that competition produced in the browser
market before Microsoft achieved dominance, and could produce in operating
systems (and other middleware) as well. Consumers feel this injury every time
another hour is lost from a crash caused by an operating system that is under
no competitive pressure to improve stability. The plaintiffs’ evidence
substantiates the breadth of consumer harm, see Henderson Dec. ¶¶ 28, 92-98;
Romer Dec., ¶¶ 4-5, 11, 14, as did SIIA’s amicus brief on liability
issues (at 32-41).[1]That
these lost benefits are not easy to quantify does not diminish their
significance. Moreover, because the antitrust laws presume harm from illegal
monopolization — only a harmful monopoly would have to act illegally to protect
its position — such quantification is irrelevant to the remedial issues here.
See Blue Cross & Blue Shield United v. Marshfield Clinic, Inc.,
152 F.3d 588, 591 (7th Cir. 1998) (Posner, C.J.), cert. denied, 525 U.S. 1071
(1999). Indeed, that a “plaintiff is unable to quantify the harm that the
defendant's practice has inflicted” supports strong injunctive relief
rather than weighing against it. Ibid.
To contend successfully that consumers are not harmed
by the monopolistic suppression of competition that has insulated Windows,
Microsoft would have to convince the Court of one of two equally merit less
propositions. First, Microsoft might claim that PC operating systems constitute
an unimportant market. But Microsoft has acknowledged elsewhere that the 130
million PC operating systems sold per year are and will remain “at the center
of” computing.[1]Second,
Microsoft might contend that competition does not provide consumer benefits in
the PC operating systems market. But that argument has been “foreclose[d]” by
Congress and the Supreme Court. National Society of Professional Engineers
v. United States, 435 U.S. 679, 690 (1978).
b. When
“a firm found to have monopoly power has committed a substantial antitrust
violation,” a remedy must “make the market more structurally
competitive.” Phillip Areeda &
Herbert Hovenkamp, Antitrust Law 207 (1999 supp.).Microsoft plainly has
“monopoly power” and Microsoft’s conduct plainly amounts to “a substantial
antitrust violation.” See Conclusions, pp. 6-7, 20-21, 24, 26, 32-34.Through a
“deliberate assault upon entrepreneurial efforts that * * * could well have
enabled the introduction of competition into the market for Intel-compatible PC
operating systems,” Microsoft “trammeled the competitive process through which
the computer software industry generally stimulates innovation and conduces to
the optimum benefit of consumers.” Id. at 20.
This is not a case of a firm with a large market share
that strayed over the legal limit in a few isolated instances. Rather,
Microsoft conducted a monopolistic campaign of a range and breadth with few
parallels in American economic history — and this record reflects only a snapshot
of a company that has made anticompetitive activity a primary business tactic
for many years. Microsoft’s effort to protect its operating systems monopoly —
and to extend that monopoly into the Internet browser market — encompassed a
full catalogue of exclusionary practices: market division proposals, coercive
exclusive arrangements, tying, and many less orthodox forms of predatory
conduct that exploited Microsoft’s market power.
In particular, the Findings are replete with
successful schemes “in which [Microsoft] has applied its monopoly power to the
task of protecting the applications barrier to entry.” Findings ¶ 340.As this
Court recognized, the core of “Microsoft’s business strategy” is to “direct[]
its monopoly power toward inducing other companies to abandon projects that
threaten Microsoft and toward punishing those companies that resist.” Id.
¶ 132.It is “Microsoft’s corporate practice to pressure other firms to halt
software development that either shows the potential to weaken the applications
barrier to entry or competes directly with Microsoft’s most cherished software
products.” Id. ¶ 93.
In this Court’s words, “Microsoft has demonstrated
that it will use its prodigious market power and immense profits to harm any
firm that insists on pursuing initiatives that could intensify competition
against one of Microsoft’s core products.” Findings ¶ 412.The evidence in this
case, and the unfortunate experience under the current consent decree, amply
demonstrate that Microsoft will abuse monopoly power by any means available to
it.
As a result, the remedy for Microsoft’s extraordinary
array of antitrust violations must go beyond narrow behavioral proscriptions
touching on particular tactics that Microsoft has used with exclusionary effect.
Rather, the remedy must deprive Microsoft of the incentive and the ability to
continue and enlarge its consistent, harmful course of conduct. So long as
Microsoft possesses its monopoly, and maintains control over the applications
barrier to entry that insulates and helps perpetuate that monopoly, nothing
will change. That is unmistakably clear both from the plaintiffs’ evidence (see
Henderson Dec. ¶¶ 6-8, 12-13, 18, 69-70, 96-98) and from Microsoft’s counterproposal
of an injunction that would forbid little, if any, of the conduct that this
Court found illegal.
c. An
ideal remedy would produce competition immediately, as the Economists’ Brief
suggests (at 8, 49-67). Ironically, Microsoft seems to endorse this approach,
criticizing the plaintiffs’ proposal because it “will not result in an
immediate and certain increase in competition in any market defined by the
Court.” Microsoft Summ. Resp. 2.If this Court agrees that relief must succeed
with certainty, then it should take Microsoft at its word and follow the
recommendation in the Economists’ Brief.
Plaintiffs’ Proposed Final Judgment takes a more
cautious course, but one that aims to correct the structural condition that now
protects Microsoft’s monopoly: the applications barrier to entry. See Findings
¶¶ 36-52.If something is not done to erode the applications barrier, there is a
clear danger that the same problems will recur again and again, in modified
forms.
Judge Greene recognized this problem when he decided
to impose a divestiture remedy in the AT&T case over the strong
objections of the Department of Defense and several States. See United
States v. AT&T, 552 F. Supp. 131, 153-154 (D.D.C. 1982) (noting
opposition by States), aff’d per curiam sub nom. Maryland v. United
States, 460 U.S. 1001 (1983); id. at 208-209 (Dep’t of Defense).[1]Judge
Greene acknowledged that “[i]t would be difficult to formulate an order that
would effectively deal with all of the different kinds of anticompetitive
behavior” at different times “and with respect to many different subjects.” Id.
at 167.That was particularly so because, in the judge’s view, the Bell System’s
“pattern” — like Microsoft’s — “has been to shift from one anticompetitive
activity to another” once one goal had been achieved or a particular practice
drew government scrutiny. Ibid.
In this case, as in AT&T, “it is unlikely
that, realistically, an injunction could be drafted that would be both
sufficiently detailed to bar specific anticompetitive conduct yet sufficiently
broad to prevent the various conceivable kinds of behavior” that an aggressive
monopolist with demonstrated contempt for the antitrust laws “might employ in
the future.”552 F. Supp. at 168.Thus, Judge Greene observed, “courts have
generally rejected this type of detailed injunction in favor of the ‘surer,
cleaner remedy of divestiture.’” 552 F. Supp. at 168 n.155 (quoting United
States v. E.I. du Pont de Nemours & Co., 366 U.S. 316, 334
(1961), and citing United States v. Paramount Pictures, Inc., 334
U.S. 131, 165-175 (1948), and United States v. Crescent Amusement Co.,
323 U.S. 173, 189-190 (1944)). The same considerations lead to the same
conclusion here.
2. Conduct
Restrictions Are Plainly Inadequate To Constrain Microsoft’s Anticompetitive
Threat Or To Reinvigorate Competition
One thing should be obvious from the breadth of the
anticompetitive conduct proved in this case and the adverse experience with the
current consent decree: conduct remedies alone will not work against a powerful
monopolist like Microsoft with no disposition to abide by the antitrust laws.
The problem in such a case — the problem in this case — is the
combination of durable monopoly power with the ability, incentive, and
inclination to misuse it. A conduct remedy will do nothing to rectify any of
these conditions.
Conduct remedies already have proved inadequate.
Microsoft has been subject to the conduct remedies in the 1995 consent decree
for five years, but the structure of the operating systems market has not
changed. To the contrary, Microsoft’s market share has grown, see Findings ¶
35, and its monopoly has become stronger with the addition of a monopoly over
Internet browsers. See Henderson Dec. ¶ 41.The current experiment with a
conduct remedy has been costly: while the 1995 decree has been in effect,
competition in the browser market was lost to Microsoft’s anticompetitive
conduct. Another loss of a competitive, innovative market that threatens the
Windows monopoly would be devastating. Not only is there no hope that conduct
remedies alone will open up the operating systems market to competition, or
even significantly restrain Microsoft from committing additional
anticompetitive acts, but conduct remedies would create insuperable enforcement
problems and heavy supervisory burdens for the Court.
The Supreme Court long has understood the inadequacy
of conduct remedies in cases involving substantial violations of Section 2 of
the Sherman Act. In a monopolization case courts must “start from the premise
that an injunction against future violations is not adequate to protect the
public interest.” Schine Chain Theatres, Inc. v. United States,
334 U.S. 110, 128 (1948). The leading antitrust treatise agrees that the
“presumptive rule” should “favor[] maximum feasible relief against monopoly
based in part on significant exclusionary practices.” 3 Areeda & Hovenkamp, supra,
¶ 655, at 101 (rev. ed. 1996).
That is because “[t]he track records of [conduct]
remedies in dislodging monopoly power have * * * not been very promising.” Areeda & Hovenkamp, supra, ¶
704.3, at 212 (1999 supp.). The reason for that failure is plain.” Simply
enjoining [particular] practices without attacking the structural monopoly does
no more than encourage the monopolist to look for some new way of exercising
its dominance that is not covered by the current injunction.” Id. at 213.
a. The
Limits Inherent in Drafting a Conduct Decree Make Evasion Inevitable.
The principal difficulty with a decree confined to
conduct relief is that the injunction must reliably prohibit every type of violation
in which the defendant has engaged, as well as foreseeable conduct that the
offender might substitute if its past methods were forbidden. In cases
involving one or two isolated practices, conduct remedies may make sense. Here,
however, where Microsoft has shown virtually limitless ingenuity in devising
ways to protect its monopoly, a conduct remedy standing alone would be doomed
to fail just as the current consent decree has failed. A decree that did no
more than catalogue, characterize, and prohibit each type of anticompetitive
conduct proved at trial would be enormously complex but easy to evade.
That is exactly what happened with the 1995 consent
decree over which this Court has presided. Microsoft has either violated — or evaded
by shifting the focus of proscribed conduct — almost every provision of that
decree. The decree prohibits Windows license agreements for longer than one
year — but only with OEMs. United States v. Microsoft Corp., No.
94-1564, 1995–2 Trade Cas. ¶ 71,096, at § IV(A) (D.D.C.) (“1995 Decree”).
Accordingly, Microsoft, directly and through non-OEM resellers, has entered
3-year licensing agreements with the business enterprises that are responsible
for most of Microsoft’s profits; those agreements include Windows upgrades for
the term of the agreement.[1]Those
agreements typically involve a lump sum for a bundle of products, based on the
number of PC desktops in an organization, rather than on actual installation of
software.[1]Thus,
were it not for the limitation of the 1995 decree restrictions to OEMs,
Microsoft’s enterprise agreements likely would also violate the decree
prohibitions on per processor licenses (§ IV(C)) and lump sum pricing (§
IV(H)), and the restriction of Microsoft Windows licenses to “per copy”
agreements (§ IV(D).
The current consent decree prohibits Microsoft from
entering into any “License Agreement” that limits an OEM’s distribution of any
non-Microsoft operating system software product.1995 Decree § IV(B). But
a “License Agreement” was defined to include any OEM license of “Covered
Product(s)” (id. § II(4)), which in turn were defined to include
stand-alone products that perform OS functions (id. § II(1)), as
Microsoft claims that Internet Explorer does. Thus, Microsoft’s coercive exclusionary
agreements aimed at Netscape likely violated this provision as well as Section
2 of the Sherman Act. And the trial evidence has shown that Microsoft
disregarded the 1995 decree’s prohibition (§ IV(E)) of OEM Windows licenses
that were “expressly or impliedly conditioned” on the licensing of any other
product or on “the OEM not licensing, purchasing, using, or distributing any
non-Microsoft product.” The tie between Internet Explorer and Windows plainly
violated this provision before Windows 98 was released, and — on the record
developed at trial — may well have violated the consent decree afterward as
well, despite the proviso permitting Microsoft to develop “integrated”
products. See United States v. Microsoft Corp., 147 F.3d 935, 950
& n.15, 952 (D.C. Cir. 1998) (Microsoft II) (limiting dictum about
“integration” of Windows and Internet Explorer to “the facts before us”).
Microsoft also evaded this provision by taking advantage of its narrow
limitation to OEMs; Microsoft’s contracts with ISPs and OLSs imposed identical
competitive injuries.
The experience with the 1995 decree shows that conduct
proscriptions alone are entirely inadequate to contain a firm as dedicated to
the destruction of competition as Microsoft. And while Microsoft’s own proposed
remedy might be dismissed as a caricature of ineffectual antitrust relief, it
nonetheless provides further illustration of the inherent weakness of conduct
restrictions. Section 4 of that proposal purports to address Microsoft’s
illegal restrictions on the boot-up sequence, icon displays, and the default
browser assignment. But it does not do that at all. To the contrary, that
provision would allow Microsoft to use discriminatory pricing and services to
coerce OEMs to forgo all of these options, or to retaliate in any other way, so
long as Microsoft did not outright cancel or refuse a Windows license
altogether. That would allow Microsoft to repeat much of the conduct condemned
by this Court. The proposed default browser provision is meaningless under its
own terms, since Section 4(b)(c) permits Microsoft to displace the “default”
browser whenever a Web page is optimized for a Microsoft browser. Section 5 of
Microsoft’s proposed decree, while barring contracts that exchange placement on
the Windows desktop for a third party’s agreement not to aid Microsoft platform
(but not applications) software, would not stop Microsoft from simply refusing
to offer such placement to parties that aided Microsoft’s competitors — much as
Microsoft did with Oakland Section 6 would do little more than give judicial
approval to Microsoft’s discriminatory treatment of ISVs, allowing Microsoft to
refuse access to technical information on any ground related to Microsoft’s
limitless conception of its own intellectual property rights. Section 7 of
Microsoft’s proposal would not stop Microsoft from repeating the coercive
conduct aimed at Apple (Findings ¶¶ 104-110; Conclusions, p.17), since that
provision would apply only to Microsoft software that was “ready for commercial
release” at the time of the threat. Although Microsoft’s proposal should not be
confused with a good-faith effort to devise an effective antitrust decree, it
nonetheless shows how superficially attractive conduct restrictions in fact
restrict nothing at all.
Even conduct restrictions devised in good faith, like
those in Plaintiffs’ Proposed Final Judgment, are valuable primarily as
supplements to structural relief — as interim measures to slow down Microsoft’s
anticompetitive activity until the reorganization takes effect. Without a more
direct approach to the problem of monopoly power, those conduct provisions at
most would make Microsoft alter some details of its practices without
materially affecting its anticompetitive strategy. For example, without a
reorganization, the conduct provisions in Section 3 of the plaintiffs’ proposal
would permit Microsoft to use identical tactics to exert its market power over
OEMs so long as the tactics involved the Microsoft Office monopoly, even if the
goal of that pressure was to preserve the Windows monopoly. Other provisions,
like the general prohibition on exclusivity, simply echo the unenforced
provisions of the 1995 consent decree, and at best would make Microsoft shift
its exclusionary tactics from one method to another. And, although the
disclosure provisions in Section 3(b) should prevent the continued misuse of
the Windows monopoly to bolster the applications barrier to entry by providing
an unfair advantage to in-house applications and middleware developers, without
a reorganization that provision would be unlikely to alter the status quo — and
would be particularly difficult to enforce.
A few obvious pathways to evasion exemplify the
inherent limitations in the conduct provisions of Plaintiffs’ Proposed Final
Judgment. Because some restrictions apply only to enumerated relationships with
OEMs, ISVs, hardware vendors, ISPs, or other third parties, Microsoft could
argue that the decree permitted it to engage in exactly the same
exclusionary behavior within the context of end-user licensing. Like other
software vendors, Microsoft already is moving toward a software-as-service
model in which at least large end-users license software for a fixed period,
including all upgrades released during that time. See, e.g., Microsoft
Summ. Resp. 17; Mary Jo Foley, Microsoft: The Next Generation, ZDNet
News, Apr. 28, 2000,
http://www.zdnet.com/zdnn/stories/news/0,4586,2556472,00.html; Mary Jo Foley, Microsoft
Dabbles in Software Hosting Waters, Sm@rt
Reseller, June 2, 1999, http://www.zdnet.com/sr/stories/0,4538,
2268932,00.html; Gary Bolles, Road to Software Rentals, Sm@rt Reseller, Mar. 8, 1999, http://www.zdnet.com/sr/stories/issue/0,4537,392493,00.html.
Though some of this software rental occurs through third-party application
service providers, it would be a simple matter for Microsoft to recharacterize
the software licenses as direct licenses to the end-users.
In
addition, Microsoft could argue that many of the conduct restrictions in
Plaintiffs’ Proposed Final Judgment apply to only one version of Microsoft’s
flagship Windows 2000 operating system, Windows 2000 Professional, which runs
single desktops and can operate as a server for small networks. See Plaintiffs’
Proposed Final Judgment § 7(dd). The other versions of Windows 2000 contain
almost exactly the same code — except that additional server features are
added, and the operating system can run on more central processing units. It
would be a simple matter for Microsoft to repackage its operating systems to
position one of the server-oriented brands as the desktop operating system as
well, and then discontinue Windows 2000 Professional. Once Microsoft’s decree
obligation to maintain the superseded operating system expired, Microsoft could
assert that, under the language of the decree, its remaining operating systems
were not subject to the conduct restraints.
These examples merely illustrate the obvious — that
conduct decrees by their nature cannot anticipate and keep pace with a
determined and devious violator.[1]Experience
with the current consent decree proves that Microsoft will find a way to work
around any remedy that does not directly and effectively address its market
power. As Microsoft has boasted, the 1995 decree did not “change [Microsoft’s]
business practices at all,” Gov’t Ex. 940, and nothing will change after
another conduct decree. See James Grimaldi, Microsoft Defends Its Practices;
CEO Ballmer Sees No Need to Change, Wash.
Post, Apr. 19, 2000.Even if this Court retains broad power to modify the
decree, the Court will not be able to respond when it discovers that the
conduct restrictions contain loopholes that permit evasion, unless it
undertakes the burden and delay of a “complete hearing and findings of fact.”
United States v. Western Electric Co., 894 F.2d 430, 438 (D.C. Cir.
1990) (quoting Brown v. Neeb, 644 F.2d 551, 560 (6th Cir. 1981));
see also id. at 434.
Microsoft’s monopoly power also allows it to continue
to mandate the adoption of Microsoft-friendly standards. The trial evidence
revealed Microsoft’s efforts to undermine standard Java because the standard
would increase cross-platform programming efficiency. The evidence also showed
that Microsoft intends to co-opt other standards, including HTML, the language
that has driven the growth of the World Wide Web, and weigh it down with
Microsoft-specific extensions. See Findings ¶¶ 233, 322; 11/10/98 (a.m.) Tr.
21-22; Gov’t Ex. 564, at 3; see also Henderson Dec. ¶¶ 34-36.This activity has
continued, as Microsoft has attempted to undermine the Kerberos security
standard. See id. ¶¶ 49-51; Romer Dec. ¶ 35-36; Felten Dec. ¶¶
77-81.This ability to dictate standards, like Microsoft’s ability to retard software
innovation that depends on interoperation with the operating system, is a
particularly dangerous aspect of illegal monopolization because it permits
Microsoft to extend its monopoly into Internet computing while protecting the
monopolies it already possesses. Microsoft’s predilection to sabotage threatening
cross-platform standards is impossible to confine through a series of
behavioral orders. An effective remedy must attack Microsoft’s repeatedly and
ingeniously abused market power, not its anticompetitive tactic du jour.
Also weighing against a conduct remedy in this case is
the need to address where Microsoft’s monopoly is going, not just where it has
been. Microsoft already has used its operating systems monopoly to breed
monopolies in the Internet browser and desktop productivity applications. A
behavioral remedy would have to be sufficient to protect competition in
adjacent markets most susceptible to future monopolization. Those include the
markets for directory services and streaming media software (both already tied
to Windows 2000), enterprise server operating systems, enterprise databases,
enterprise e-mail and collaborative computing servers, enterprise server
applications, Internet infrastructure software (web servers, proxy servers, applications
servers), consumer finance and transaction applications, and online services
providing proprietary content. Microsoft already has used its ability to
insinuate technological links between its desktop monopolies and enterprise
software, and to bundle its Windows operating systems with other software (both
technologically and contractually), to accelerate its acquisition of market
power in enterprise software markets. See Henderson Dec. ¶¶ 38-40; Romer Dec.
¶¶ 35-36; see also SIIA Br. Supporting U.S. 9-13.Indeed, 40% of the
functionality of the desktop version of Windows 2000 is useless without a
Windows 2000 server. See Ballmer Is Bullish on Windows 2000,PC Week, Nov. 19, 1999,
http://www.zdnet.com/pcweek/stories/news/0,4153,1018247,00.html.
Microsoft also has hinted that the strategy of its
“Next Generation Windows Services” is to insert “technological shackles”
(Conclusions, p.10) in its monopoly desktop software to compel the use of
Microsoft software throughout the Internet, so that the ocean of innovation on
the Internet becomes a stagnant Microsoft-proprietary pond. Dominic Gates, How
to Integrate Everything: Undaunted by its Antitrust Defeat, Microsoft Plows
Ahead with its Plan to Dominate the Internet, Industry Standard, May 22, 2000, at 120; see id. at
121 (noting that reorganization proposed by plaintiffs in this case likely
would prevent that result); see David Kirkpatrick, The New Face of Microsoft, Fortune, Feb. 7, 2000, at 87, 88
(Microsoft is aiming for “a new kind of operating system, in which Windows gets
spread throughout the Internet, just as the current Windows is in every nook
and cranny of your computer’s hard drive”). See also Microsoft Summ. Resp.
17-18.Such conduct is impossible to predict, much less prohibit in advance with
enforceable specificity.
b. Effective
and Timely Enforcement of Conduct Restrictions Is Practically Impossible.
Even if it were possible to craft a set of behavioral
remedies that could not profitably be circumvented, it would be necessary to
ensure that Microsoft complied with those restrictions. Effective monitoring
and enforcement would require continuing oversight of Microsoft’s commercial
activities by government agencies and the court. As the Supreme Court has
observed, “the policing of an injunction would probably involve the courts and
the Government in regulation of private affairs more deeply than the
administration of a simple order of divestiture.” Du Pont, 366 U.S. at
334.Accord, AT&T, 552 F. Supp. at 167-168.Comprehensive behavioral
decrees inevitably require interpretation and application as the defendant
introduces new products, moves into new markets, or changes its business strategies
in its traditional markets. Microsoft and the government plaintiffs would
return to court repeatedly to argue over the boundaries of the decree
prohibitions and to determine their application to Microsoft’s latest
exclusionary initiatives. Given Microsoft’s litigious bent and enormous wealth,
this would tax the scarce resources of the Court and the enforcement
authorities to the breaking point.
One
of Microsoft’s leading legal public relations spokesmen, Charles F. (Rick)
Rule, has explained the contrast between the efficiency of structural relief
and the comparative inefficiency of behavioral relief in the AT&T
case. Rule inherited the Bell System divestiture decree and had to live with it
as Assistant Attorney General for Antitrust in the 1980s.Near the end of his
tenure in that position, Rule gave an accounting of “the lessons of the
AT&T decree.” See Antitrust and Bottleneck Monopolies: The Lessons of
the AT&T Decree, Remarks of Charles F. Rule before The Brookings
Institution, Developments in Telecommunications Policy, Oct. 5, 1988 (“Rule
Speech”).
Rule
praised the divestiture: “there is little doubt that the breakup of AT&T
achieved substantial benefits.” Rule Speech 14.But continued supervision of the
industry under the conduct remedies imposed in the decree created an enormous
problem.” Unfortunately, the choice of the Department [of Justice] and decree
court as [regulators] was, in my opinion, a big mistake.” Id. at 21.In
Rule’s words, “Unless you have served on the front lines of the Department’s
peacekeeping force, it is hard to imagine the barrage of time-consuming and
often petty complaints that constantly bombards the Department.” Id. at
23.
In
this case, federal and state enforcement agencies would have to monitor the
day-to-day operations of a fast-moving business with 35,000 employees, many
products, and frequent upgrades and new product introductions. They would have
to investigate complaints of non-compliance, debate decree interpretation and
modification issues, and litigate contempt proceedings. The difficulty of this
task would be most acute regarding decree provisions that attempted to restrict
Microsoft’s decisions regarding compatibility or other technical matters, or
the degree to which information may flow between divisions of the firm. These
would be the hardest for outsiders to observe and the easiest for Microsoft to
explain away with arguments that the actions did not fall squarely within a
particular conduct restriction.
Those
drawbacks would be magnified given Microsoft’s evident disregard of the legal
norms imposed by antitrust laws. Microsoft responded to this Court’s effort to
enforce the anti-tying provisions of the 1995 consent decree by proposing to
distribute a version of Windows that did not work — and with a straight face
told this Court that its order required such a response. See Microsoft
Asserts Compliance With Injunction in Only Way Possible, 74 Antitrust & Trade Reg. Rep. (BNA)
49 (Jan. 15, 1998).
Not
only did Microsoft successfully dodge the 1995 consent decree, but its conduct
related to this case makes clear that a conduct remedy would be a wasted
effort. Microsoft executives at several junctures appeared to be incapable of
simply telling the truth under oath and, during James Allchin’s trial
testimony, Microsoft presented this Court with a crudely faked videotape. Even
if the Court concludes that these incidents do not warrant reference to the
United States Attorney for criminal investigation, they certainly indicate that
Microsoft does not take legal obligations seriously and cannot be trusted to
comply with them. Nothing would change under a remedy that required only
compliance with specific conduct proscriptions — except that Microsoft might
refine its deceptions so as to escape detection. Meanwhile, outside the
courtroom, Microsoft has campaigned for legislation to undermine antitrust
enforcement, while trying to intimidate state and federal law enforcement
officials with a massive lobbying and public relations campaign that equates a
reduction in Microsoft’s monopoly rents with damage to the public welfare.[1]Success
in that undertaking would, of course, render any conduct decree meaningless.
Microsoft merely confirmed its contempt for the
antitrust laws and this enforcement proceeding when it proposed a remedy that
would not prohibit most of the conduct at issue in this case, and would permit
the rest with slight variations. Indeed, in its proposal Microsoft continues to
refuse to implement an antitrust compliance program. See 1/27/99 (a.m.) Tr. 5
(Maritz) (Microsoft has no antitrust compliance program); see also United
States v. Microsoft Corp., 159 F.R.D. 318, 336-337 (D.D.C.) (noting
Microsoft’s refusal to implement antitrust compliance program), rev’d, 56 F.3d
1448 (D.C. Cir. 1995); but see Plaintiffs’ Proposed Final Judgment § 4
(requiring such a program). Rather than take steps to ensure compliance with
the antitrust laws, Microsoft created a new high-level management position to
focus on the public relations aspects of its coercive and exclusionary conduct.
See Kara Swisher, Microsoft Appointee: Change Agent or PR Ploy?, Wall St. J., May 15, 2000, at B1.
Microsoft’s refusal to bring its market conduct within
legal bounds is chronic and intractable. While this Court has been considering
Microsoft’s liability, Microsoft has undertaken to redesign its software to
injure Palm, a competitor that it views as posing a long-term threat to the
Windows monopoly. See Plaintiffs’ Mem. Supp. Proposed Final Judgment 10 (citing
Gov’t Rem. Exs. 1-2); Henderson Dec. ¶ 69. See also Henderson Dec. ¶ 44
(explaining current need for handheld devices to interoperate with PCs). And in
its efforts to foreclose another middleware threat, Microsoft has duplicated
the “technological shackl[ing]” that is one of the key elements of the conduct
found illegal in this case (Conclusions, p.10). To avoid having to compete on
the merits in the market for streaming media, Microsoft is duplicating the
tying arrangement it used to gain a monopoly over Internet browsers. See Microsoft
Unleashes New Media Player, May 2, 2000,
http://www.zdnet.com/zdnn/stories/news/0,4586,2560616,00.html (Reuters story)
(noting that the Microsoft media player, unlike others, can be used only in
conjunction with Microsoft media server software). Microsoft considered
RealNetworks “an OS contender” (Gov’t Ex. 1368), and this Court explicitly
recognized the middleware threat posed by streaming media. See Findings ¶¶
104-114.
Even now, Bill Gates continues to declare that
“Microsoft is very clear that it has done absolutely nothing wrong,” Steve Lohr
& Joel Brinkley, Microsoft Management Tells Workers There Will Be No
Breakup, N.Y. Times,
Apr. 26, 2000, at C1, C9, an attitude reflected in its proposed remedy, which seems
to assume that Microsoft won this case rather than lost it. Microsoft did not
“change [its] business practices at all” after the 1995 consent decree. See
Gov’t Ex. 940.It “ha[s]n’t changed anything” in response to the current case.
Henderson Dec. ¶ 97 (quoting Brad Chase from Gov’t Rem. Ex. 17); see also
Grimaldi, Microsoft Defends Its Practices, supra, at E1
(paraphrasing Steve Ballmer remark that Microsoft “had no reason to change its
business practices”). And, if its proposed remedy were all that resulted from
this case, Microsoft would not change anything after judgment was entered in
this case despite the sweeping liability that this Court found. Given
Microsoft’s corporate history and inability to distinguish right from wrong
under the antitrust laws, as confirmed by the proposed remedy it filed with
this Court, “an injunction can hardly be detailed enough to cover in advance
all the many fashions in which improper influence might manifest itself.” Du
Pont, 366 U.S. at 334.
Beyond this, the inevitable delay between
anticompetitive conduct and a final order enforcing a decree vastly reduces the
effectiveness of even an accurate and comprehensive “good conduct” regime.
Nothing would happen to Microsoft upon commission of an anticompetitive act; probably
nothing would happen before the act had the full desired effect in the
marketplace. See Henderson Dec. ¶ 111 (quoting Gov’t Ex. 1458, at 209). Much
time would elapse before Microsoft had to comply with a contempt order, if any
order could be secured. See Microsoft II, 147 F.3d at 940 (conduct
prohibitions that are ambiguous as applied to particular business practices
cannot be enforced by contempt). The act would have to be detected by rivals or
customers, then reported to the enforcement agencies, then investigated and
evaluated by those agencies, then submitted to this Court, whose order no doubt
would be appealed and might be stayed until the appeal was decided.
That is exactly what happened under the 1995 consent
decree. By the time the government had gathered enough evidence to bring a
case, Microsoft was only a few months away from releasing Windows 98.When this
Court enjoined the release of that product according to Microsoft’s plans,
Microsoft argued that the timely release of its next monopoly operating system
was critical to the health of the entire economy. Not only was the injunction
stayed pending appeal, but the court of appeals held that, given ambiguities in
the decree, this Court should have conducted a more comprehensive proceeding. That
is, when decree terms are not perfectly clear — a common pitfall when complex
technology is at issue — the enforcement process will take even longer.
This enforcement lag alone provides a compelling
reason not to make conduct restrictions the centerpiece of a decree, rather
than a supplement to a structural solution.[1]Although
the plaintiffs might haul Microsoft back into court for a contempt hearing or
decree modification, Microsoft likely would avoid a preliminary injunction or
other swift relief in all but the most obvious cases. Microsoft could continue
to squeeze out competition while this Court sorted out the dispute.
c. Structural
Relief Should Not Be Postponed for a Probationary Period.
For similar reasons, there is nothing to be gained by
withholding an order of structural relief until some later date, after conduct
remedies have failed. To begin with, conduct remedies have failed already.
Microsoft has been on probation for the five years since the 1995 consent
decree was entered. If that decree had contained a “crown jewels” provision
authorizing the imposition of structural relief upon proof of decree
violations, much of the current case would have been just such a proceeding.
Even if the use of “technological shackles” (Conclusions, p. 10) to tie
Internet Explorer to Windows perhaps did not violate the 1995 decree, the
contractual tying and coercive exclusivity proved in this case almost certainly
did. See 1995 Decree § E(i), (ii).
To give Microsoft another probationary period
not only would be an undeserved windfall, but would require an entire new trial
before significant relief could be imposed — a trial for which the plaintiffs would
have even greater difficulty than before finding witnesses who would risk
Microsoft’s retaliation when two prior cases had produced nothing but
admonitions against bad behavior. The likely — if not inevitable — need for an
additional full-scale proceeding contrasts sharply with the “surer, cleaner
remedy” of reorganization that has earned the Supreme Court’s endorsement. Du
Pont, 366 U.S. at 334.
Moreover, if consideration of structural relief is
postponed, any order actually imposing such relief would be appealed and
delayed still further, while an order imposing structural relief now can be
passed upon by higher courts without delay. Delaying structural relief would
spawn years of uncertainty for customers, investors, and employees, all of whom
would know that Microsoft’s conduct could trigger a new divestiture proceeding
at any time.
There is no danger of precipitous remedial action if
Plaintiffs’ Proposed Final Judgment is entered now. The actual reorganization will
not take place until the Supreme Court affirms this Court’s judgment and
remands the case. If, by the time of the remand, circumstances have changed so
dramatically that no reorganization is warranted — if, that is, if Microsoft no
longer has monopoly power — then this Court will be able to modify the remedy
accordingly. See, e.g., United States v. Aluminum Co., 148
F.2d 416, 446-447 (2d Cir. 1945) (finding that divestiture would be
inappropriate if federal government succeeded in disposing, to suitable
competitors, of manufacturing capacity in monopolized market built and owned
by federal government during World War II).
3. The
Law of Antitrust Remedies Compels Structural Relief For Pervasive
Monopolization Offenses
a. Under
established antitrust remedial principles, a structural reorganization is the
preferred remedy in a major monopolization case brought by the government. This
is the biggest and most egregious Section 2 violation that the government has
litigated to judgment since Standard Oil, one matched in scope in the
interim only by the partially litigated AT&T case. A governmental
remedy should “pry open to competition a market that has been closed by [a]
defendant[’s] illegal restraints.” Ford Motor Co. v. United States,
405 U.S. 562, 577-578 (1972). To do that, a remedy cannot merely nibble around
the edges of the monopolized market, but should make sufficiently fundamental
changes to allow competition on the merits. The increased efficacy and reduced
reliance on judicial oversight characteristic of structural relief have led the
Supreme Court to recognize divestiture as “the most important of antitrust
remedies.” California v. American Stores Co., 495 U.S. 271, 281
(1990) (quoting Du Pont, 366 U.S. at 330-331). Indeed, because
divestiture “is simple, relatively easy to administer, and sure,” ibid,
“the public” is ordinarily “entitled to” that “surer, cleaner
remedy” rather than one that permits a monopolist to continue as it was,
subject only to behavioral restrictions. Du Pont, 366 U.S. at 334
(emphasis added). Unless and until the Supreme Court changes the governing law,
structural relief is required in this situation.[1]
The remedy is the most important part of an antitrust
case, see Glaxo Group, 410 U.S. at 64, because a mere declaration of
wrongdoing, if unaccompanied by tangible consequences, makes government
antitrust litigation “a futile exercise.” Du Pont, 366 U.S. at
323.Relief in an antitrust case should both “cure the ill effects of the
illegal conduct, and assure the public freedom from its continuance.” Ibid.
(quoting United States v. United States Gypsum, 340 U.S. 76, 88
(1950)). When non-competitive market structure has permitted the “illegal
conduct” to succeed, the remedy must address that structure. See Areeda & Hovenkamp, supra,
at 207 (1999 supp.).
Protestations that structural relief goes too far in a
monopolization case repeatedly have fallen on deaf ears in the Supreme Court
and should meet the same fate here. A reorganization is not a disproportionate
remedy for monopolization, but rather is the only appropriate remedy for a
substantial monopolization violation like that proved in this case. See Areeda & Hovenkamp, supra,
at 207 (1999 supp.); United Shoe, 391 U.S. at 250.The Supreme Court has
rejected contentions “that the injunction should go no farther than the
violation or threat of violation,” International Salt Co. v. United
States, 332 U.S. 392, 400 (1947), and that “antitrust violators may not be
required to do more than return the market to the status quo ante.” Ford,
405 U.S. at 573 n.8.The “consequences of proved violations” extend beyond the
precise scope of the violations themselves, International Salt, 332 U.S.
at 400, and certainly do here. As a result, the Supreme Court has characterized
as “absurd” the suggestion that a monopolist that has “restrain[ed] commerce by
suppressing competition * * * must be left in possession of the power that it
has acquired, with full freedom to exercise it.” Ford, 405 U.S. at 574
n.9 (quoting Northern Securities Co. v. United States, 193 U.S.
197, 357 (1904)). To the contrary, a decree should “render impotent the
monopoly power found to be in violation” of the antitrust laws. United
States v. Grinnell, 384 U.S. 563, 577 (1966). And, in formulating
relief, “any plausible doubts should be resolved against the monopolist.” 3 Areeda & Hovenkamp, supra,
¶ 653c1, at 96.
In addition, because the relief is in the form of an injunction,
the remedy necessarily will be forward-looking. A remedy should take into
account “probable future trends in the * * * market” that are “visible at the
time” of the violation, Ford, 405 U.S. at 580 (Stewart, J., concurring
in the judgment), because “Congress * * * intended” antitrust “decrees to deal
with the future economic condition of the enterprise as well as past
violations. “International Salt, 332 U.S. at 401 n.10.Thus, a remedy’s
effectiveness should be judged with respect to where the market is going, not
where it has been. A remedy is useless if it would have made sense a few years
ago but will not alter the status quo a few years from now.
Having been “caught violating the [Sherman] Act,”
Microsoft “must expect some fencing in. “Otter Tail Power Co. v. United
States, 410 U.S. 366, 381 (1973) (quoting FTC v. National Lead
Co., 352 U.S. 419, 431 (1957)). Microsoft’s demonstrated “proclivity for
predatory practices” requires full and reliable prophylactic relief. Ibid.
b. In
its most comprehensive exposition of the standards for appropriate relief in
monopolization cases, the Supreme Court announced clear criteria. Having found
monopolization, the Court held, the decree court has a “duty * * * to
prescribe relief” that accomplishes three goals. United States v. United
Shoe Machinery Corp., 391 U.S. 244, 250 (1968) (emphasis added). First, the
relief should “terminate the illegal monopoly.”Ibid. Second, the decree
should prevent “practices likely to result in monopolization in the future.” Ibid.
Third, the order should “deny to the defendant the fruits of its statutory
violation. “Ibid. This last factor is important. The Supreme Court has
repeatedly emphasized the need to prevent a monopolist from retaining the
accrued competitive benefits of its illegal conduct. See ibid.
(collecting cases); see also Glaxo Group, 410 U.S. at 64 (quoting U.S.
Gypsum, 340 U.S. at 88). These advantages may permit a monopolist to
maintain its monopoly without additional violations of the antitrust laws. Although
a full divestiture of the Microsoft monopoly to create immediate competition in
PC operating systems (see Economists’ Brief 49-67) may be the most direct and
foolproof way to achieve all of these goals, the limited remedy set forth in
Plaintiffs’ Proposed Final Judgment serves these goals effectively.
First, although Plaintiffs’ Proposed Final Judgment
would not immediately “terminate the illegal” operating systems monopoly, United
Shoe, 391 U.S. at 250, it would deprive Microsoft of two of the biggest
barriers now preventing the emergence of a competitive operating systems
market. To begin with, by separating the browser development assets from the
operating systems assets, the proposed reorganization reinvigorates middleware
as a competitive threat to Windows. After the reorganization, Internet
Explorer may continue to be more closely integrated with Microsoft applications
than with competing applications, but that will not necessarily tie users to
Windows, particularly if Office is ported to other operating systems. In
addition, by separating Windows from Office, plaintiffs’ proposal erodes the
most important element of the applications barrier to entry that was exploited
in this case. That erosion has two procompetitive consequences. It makes up for
the success of Microsoft’s illegal conduct in delaying or eliminating other
challenges to the operating systems monopoly. See Findings ¶¶ 411-412.It also
weakens the weapon — monopoly power — that has too often tempted Microsoft to
engage in anticompetitive abuses. The proposed reorganization attacks the
operating systems monopoly far more directly than conduct restrictions could,
while creating the possibility of renewed competition in the Internet browser
market that Microsoft has illegally monopolized during the course of this
litigation.
Second, Plaintiffs’ Proposed Final Judgment also helps
prevent the recurrence of anticompetitive practices similar to those proved at
trial and “likely to result in monopolization in the future. “United Shoe,
391 U.S. at 250.The central problem proved in this case is Microsoft’s broad,
strategic, and wide-ranging abuse of market power. The separation of the
browser and the Office suite from Windows undermines the Windows monopoly and
thus addresses the disease rather than the mere symptoms of Microsoft’s abuses.
For example, if a unitary Microsoft remained in control of Internet Explorer,
the browser could provide a ready means of further insulating the operating
systems monopoly rather than providing a potential challenge to it — as shown
by the events examined at trial. If the operating systems monopoly had to rely
on outsiders for advanced Internet browsing capability even temporarily,
however, a competitive browser market would afford many other market participants
an opportunity to sidestep Microsoft’s monopoly by orienting themselves more
toward independent, competitive browsers and less toward the Windows operating
system.
Third, Plaintiffs’ Proposed Final Judgment deprives
Microsoft of some of “the fruits” of its illegal conduct. United Shoe,
391 U.S. at 250. Much of the conduct in this case was directed at
obtaining control over Internet browsing in order to protect the Windows
monopoly. A principal benefit of that conduct was Microsoft’s success in controlling
the browser market to serve the goal of making the Internet “Windows centric.”
Gov’t Ex. 611.Plaintiffs’ Proposed Final Judgment appropriately takes that
control away from the Windows monopoly and gives it to a competing applications
company. Likewise, Microsoft tried to transform the browser from a middleware
threat to the Windows monopoly into a vehicle for extension of that
monopoly — and, indeed, made the browser into another aspect of the applications
barrier to entry. The proposed reorganization appropriately strips away the
portion of that barrier that has been added by the new Internet Explorer
monopoly. That is not “punishment”; rather, it precisely remedies a competitive
wrong.
c. Microsoft
claims that structural relief cannot be ordered against a Section 2 violator
unless the evidence clearly demonstrates that the particular competition
suppressed would have displaced the monopoly. E.g., Mem. Supp. Judgment
2; Mem. Supp. Rejection 6-8.But assessment of a market in the absence of actual
illegal conduct necessarily involves counterfactual estimation, so causation is
inherently difficult to show. Thus, it is not surprising that this Court could
not find with confidence that Navigator and Java “already would have
ignited genuine competition in the market for Intel-compatible operating
systems. “Findings ¶ 411 (emphasis added).
Microsoft’s argument proves far too much, however,
because it would give a free ride to monopolists that successfully but
illegally crushed incipient competitive threats. If structural relief cannot be
ordered in such a case, even though conduct relief would be ineffective, then
no preventive measure would stand between a monopolist like Microsoft and its
ability to abuse monopoly power to eliminate competition — at least so long as
competition is crushed while the threat is incipient. Microsoft in effect asks
to be rewarded for its own ruthless efficiency in deploying anticompetitive
conduct against potential challenges to its monopoly as soon as it recognized
those challenges. Microsoft’s self-serving argument is plainly insufficient to
overcome the presumption that structural relief is required. See Areeda & Hovenkamp, supra,
at 207 (1999 supp.). Microsoft destroyed the threats that existed; its success
in doing so before the threats were strong enough to achieve demonstrable
success should not entitle Microsoft to preserve its monopoly at the expense of
the next set of potential competitors. Cf. American Tobacco Co. v. United
States, 328 U.S. 781, 809-810 (1946) (“actual exclusion of competitors” not
necessary even in a criminal case brought under Section 2). Microsoft
also should not be heard to claim (Mem. Supp. Rejection 14-15) that it was
uncertain about the legality of its use of exclusionary conduct to protect its
monopoly. Judge Bork demonstrated in his earlier amicus brief that
liability rested squarely on established Supreme Court antitrust precedent.
Microsoft was not “uncertain” about the antitrust consequences of its action;
it simply didn’t care. Despite operating under an antitrust consent decree,
Microsoft had no antitrust compliance program to inculcate a corporate
conscience in such matters.
d. Finally,
the fact that this is a public antitrust case fundamentally affects the
remedial calculation. After government plaintiffs have sustained the burden of
proof of a major violation, they are entitled to reasonable deference in their
selection of an effective remedy. Ford, 405 U.S. at 575.Effective
structural remedies call for predictions about the future and require law
enforcement experience and expertise. Although courts must independently review
proposed remedies, they should defer to reasoned judgments of federal and state
law enforcement officials on what is feasible and effective. See United
States v. Western Electric Co., 900 F.2d 283, 294 n.12 (D.C. Cir.
1990) (referring to the Department of Justice as the “Prime Mover” of a
divestiture case).[1]The
extensive analysis offered by the plaintiffs and their witnesses amply warrants
such relief.
B. Plaintiffs’
Proposed Final Judgment Provides An Appropriate Framework For Relief
1. Reorganization
Of Microsoft Into An Applications Business And An Independent Operating Systems
Business Comports With The Liability Findings
Plaintiffs’ Proposed Final Judgment appropriately
addresses the violations found in this case: the destruction of competition in
a software application (the Internet browser) in order to insulate the
operating systems monopoly from competition; the attempted (and now completed)
acquisition of a monopoly over the Internet browser market; and the
multifaceted abuse of the operating systems monopoly that buttressed the
applications barrier to entry. Microsoft’s anticompetitive activity has relied
upon the applications barrier that protects its operating systems monopoly, and
has reinforced that barrier through extraordinary effort and ingenuity.
The Internet browser posed a threat that applications
might no longer depend on the operating system to the extent they had in the
past. Applications instead might be written to interface with the browser or
other middleware, middleware that in turn could run atop a variety of operating
systems. As Microsoft recognized, such a development could “commoditize the
underlying operating system,” Gov’t Ex. 20, returning Microsoft to a
competitive operating systems market in which it would have to price at
competitive levels rather than extracting an estimated 80% profit margin, see
Kirkpatrick, supra, at 90.That prospect fills Microsoft officials with
understandable dread. See James Grimaldi, U.S., States Favor Plan to Split
Up Microsoft, Wash. Post,
April 24, 2000, at A1, A8 (quoting Ballmer as complaining that no software
company could make money “selling the same thing” as another firm). Even more
significantly, however, a breach in the applications barrier would mean that
Microsoft no longer controlled a critical chokepoint in the software industry,
and thus would render it unable to use its power to coerce OEMs, ISVs, ISPs,
and others in the computer and software industry to keep Microsoft’s core
products free from competition.
The reorganization aims directly at the applications
barrier to entry. Separating Windows from a good portion of the applications
barrier does not exceed the scope of this case simply because the applications
barrier itself involves software products that are not operating systems. The
applications barrier to entry is the structural condition that under girds the
monopoly power that Microsoft has repeatedly abused. The applications barrier
to entry is particularly severe because the most widespread applications,
personal productivity applications, are controlled by Microsoft through its
Office suite, which accounts for almost all such suites sold for personal
computers. See Romer Dec. ¶ 17.Moreover, by adding the Internet browser to the
barrier, Microsoft’s conduct in this case successfully reinforced that
structural limitation on competition. That state of affairs makes the
reorganization more appropriate, not less so. As Judge Bork recently explained,
“Microsoft’s predation in the browser market was merely the most prominent of
its efforts to preserve the applications barrier to competition with its
operating system.” Robert Bork, There’s No Choice: Dismember Microsoft, Wall. St. J., May 1, 2000, at A34.Just
as Microsoft sought to reinforce the applications barrier to entry by
controlling the Internet browser market, through Office Microsoft controls the
most important aspect of that barrier. One monopoly thus reinforces the other.
Section1 of Plaintiffs’ Proposed Final Judgment
removes this leverage favoring the Windows monopoly — including the leverage
provided by Microsoft’s dominance of the Internet browser — by forcing
Microsoft to reorganize into one separate company for the “Operating Systems
Business” (i.e., the “Windows Company”), and another for the
“Applications Business” (i.e., the “Applications Company”). Plaintiffs’
Proposed Final Judgment does not immediately terminate the Windows monopoly,
but substantially undermines it. Rather than dictating a result or choosing a
competitive technology, Plaintiffs’ Proposed Final Judgment merely tries to
create the conditions that will permit the market to make the choice without
the distortions caused by Microsoft’s abuse of monopoly power on multiple
fronts.
To begin with, Plaintiffs’ Proposed Final Judgment
would end Microsoft’s ability to leverage the Windows monopoly into other
markets — especially middleware markets — at least until the Windows Company
could develop middleware of its own (or could acquire a middleware product
without violating the antitrust laws). The current Microsoft applications
assets, including middleware, would be owned and controlled by an Applications
Company that at the very least would compete with any middleware developed by
the Windows Company, and that would provide significant resistance to any
anticompetitive forays.
More fundamentally, the Windows Company and the
Applications Company no longer would have the incentive and ability to
reinforce each other’s monopolies. The Windows Company would lack the incentive
and ability to give Microsoft applications developers the advantage of
preferential access not only to the specifications, but to the design of the
monopoly operating system. Moreover, the Windows Company and the Applications
Company each would have strong incentives to offset and undermine the market
power of the other.
That countervailing power could take many forms because,
unlike the AT&T decree, Plaintiffs’ Proposed Final Judgment does not
place any limit on the lines of business in which the successor companies may
engage. As a consequence, each successor company could invade the other’s
market directly using its substantial capital and engineering resources: the
Windows Company could develop applications, and the Applications Company could
develop a new operating system. Short of that — and immediately — each would
have strong incentives to cooperate with the rivals of the other. No longer
tied to a corporate objective of maintaining and expanding the Windows
monopoly, the Applications Company would have an incentive to port its products
to other platforms (and to bring Office for Macintosh up to parity with Office
for Windows). That course of action would quickly undermine important elements
of the applications barrier preventing significant entry into the operating
systems market, and thus would advance the development of Linux and other
operating systems as competitive alternatives to Windows.[1]See
Shapiro Dec. § III (A), at 9-10; Henderson Dec. ¶¶ 102-106.
Even if the Applications Company did not itself port
the Office suite and other applications to non-Windows operating systems, however,
that company would have strong incentives to assist middleware developers
attempting to make Microsoft Office run smoothly on Linux and other operating
systems. The same motivation would apply across the board, to both desktop and
server applications and to a wide variety of middleware with the potential to
displace the dominance of the operating system as a platform. Likewise, the
Windows Company no longer would have the ability and incentive to aid the
Applications Company at the expense of competing applications developers. If
applications developers competed on a level playing field, that could only
increase competition — and decrease prices — for the applications that
consumers use most.
Microsoft claims that the proposed reorganization
would dissipate efficiencies that purportedly result from single-firm
ownership of the Windows monopoly, the Office monopoly, and the browser
monopoly. E.g., Summ. Resp. 11-23; Mem. Supp. Rejection 24. Those
arguments essentially restate the position Microsoft took with RealNetworks:
that “[O]ffice and Windows [a]re one, * * * and Word/Excel ([O]ffice) [a]re
part of the OS.” Gov’t Ex. 1368.But the supposed efficiencies resulting from
single-firm control over several tiers of the software industry are illusory.
The rest of the software industry (and the computer and communications
industries) long have thrived in areas where no such control— much less
monopoly control — rests in one company. This non-monopolistic market structure
has produced common standards such as TCP/IP and HTML, and has permitted
interoperability for a stunning array of products. Microsoft’s claims here —
that every software product that Microsoft decides to monopolize becomes an
integral part of the operating system as soon as Microsoft says so, see, e.g.,
Summ. Resp. 22 (identifying the Internet browser, the Web server, and streaming
media are “core Windows technologies”) — plainly demonstrate the need for
relief that both cabins and undermines this illegally maintained monopoly.
There is no technical advantage to single-firm
hegemony over multiple tiers as opposed to the disclosure of interfaces needed
to permit interoperability between products made by different software vendors.
See Felten Dec. ¶¶ 46-47; Shapiro Dec. § III (D), at 13-14.Companies can
achieve full integration of separate programs in this way. Even Microsoft had
to admit that Caldera achieved “OS Integrated Browsing” although the operating
system and the browser were not developed by the same organization. See Felten
Dec. ¶ 47 (citing Gov’t Ex. 1707 (Microsoft videotape demonstration)); 2/1/99
(p.m.) Tr. 73 (Allchin)).
Microsoft’s claim (Mem. Supp. Rejection 16) that
consumers would not have gotten the benefits of toolbars if Microsoft had not
owned both Excel and Windows is fatuous. To begin with, it is not true that
“[t]he concept of toolbars * * * was first developed by the team building
Microsoft Excel” (Summ. Resp. 20); toolbars appeared on Apple’s MacPaint 1.5
program years before Excel was first released. Moreover, Microsoft claims that
it adds improvements to Windows based on suggestions by ISVs. See Shapiro Dec.
§ III (D), at 13-14 & nn.22-23 (citing sources). In any event, no developer
could long have kept the idea of toolbars to itself; the basic toolbar functionality
would be no more protectable than the Macintosh user interface that Microsoft
copied. See Apple Computer, Inc. v. Microsoft Corp., 35 F.3d 1435
(9th Cir. 1994). Likewise, tablet PCs and voice recognition software have
emerged from companies that do not have complementary monopolies. See
Jay Greene, Don’t Worry, Bill — Innovation Will Survive, Bus. Wk., May 22, 2000, at 48.These
products will reach consumers without being offered by an operating systems
monopolist.
The main result from having Windows and Office
developers under one roof is the creation of an anticompetitive advantage:
Microsoft’s ability to engage in technological tying of competitive
applications to the operating system, and to give its own developers
preferential access to Windows functionality before competitors. Microsoft’s
propensity to intermingle applications code with the code for the monopoly
operating system increases the security dangers of a software monoculture.
Microsoft’s e-mail client, Outlook, and its word processor, Word, provide virus
writers an easy medium to attack not only those programs but also the entire
operating system. See n.1, supra.
When it was in the process of acquiring the Office
monopoly in the 1980s and early 1990s, Microsoft denied that it gave its
applications developers special access to the operating systems development
process, going so far as to claim that the “Chinese wall” between applications
developers and OS developers was “like the separation of church and state.”
James Gleick, Making Microsoft Safe for Capitalism, N.Y. Times Mag., Nov. 5, 1995, at 50, 57
(quoting Ballmer). See United States v. Microsoft Corp., 159
F.R.D. 318, 337 n.36 (D.D.C.) (noting press reports of similar claims by
Microsoft), rev’d, 56 F.3d 1448 (D.C. Cir. 1995). But see ibid. (noting
that Microsoft denied that any such separation ever existed). Bill Gates now
claims that “if we did not have Office and Windows and those groups working
together * * *, there would be no Windows as we know it.” James Grimaldi, Clinton’s
Aides Get Briefing on Microsoft, Wash.
Post, Apr. 26, 2000, at E1.“Windows as we know it” now, of course, is an
anticompetitive tool, which uses lack of operating systems choice to force
Microsoft products throughout an enterprise. See Henderson Dec. ¶¶ 39,
49-51.Beyond this, Bill Gates has confirmed what competing applications vendors
long have suspected — that independent applications developers not only do not
enjoy parity with Microsoft applications developers, but rather that “Office
and Windows * * * groups work[] together” to ensure that the products of each
reinforce the monopoly of the other.
Nonetheless, Microsoft has argued that it provides
outside applications developers full and equal access to the Windows platform,
and that Windows developers can incorporate suggestions from outside
applications developers. See Shapiro Dec. § III (D), at 13-14 & nn.22-23
(citing sources). The reorganization proposed in Plaintiffs’ Proposed Final
Judgment would make Microsoft live up to its claims about the “openness” of the
Windows platform.[1]That
is not an efficiency loss, but rather represents a gain to consumers when the
Windows and Office monopolies each have incentives to cooperate with the
competitors of the other, fostering innovation in both operating systems and
applications. As leading industry analysts recently observed, a reorganization
along the lines of Plaintiffs’ Proposed Final Judgment is “the only effective
way to prevent future wasted resources, to encourage technology innovation, and
to focus Microsoft on producing the best possible products.” Anthony Picardi
& Dan Kusnetzsky, Microsoft on Trial: An Analysis of Possible Outcomes,
International Data Corp., at 7 (1999) (“IDC Microsoft on Trial”).
It is true that, under Plaintiffs’ Proposed Final
Judgment (without the supplement proposed below), the Applications Company
would be able to try to use the monopoly Internet browser to extend the scope
of the monopoly productivity suite into server computers and Internet
computing. But there is a fundamental difference between the ability to combine
an operating system with a browser and the combination of a browser with other
applications. Every PC must have an operating system, making the operating
system monopoly a particularly dangerous and pervasive threat to competition.
There is good reason to believe that competition in applications can revive,
however, as software developers provide end-users with more ability to perform
more functions using PC resources in conjunction with Web-based programs
through the Internet, or server-based programs through an intranet. When the
distribution of productivity applications is divorced from the distribution of
the monopoly operating system, competition in productivity applications will
have an opportunity to develop – as will browser competition.
Plaintiffs’ Proposed Final Judgment poses no risks to
innovation. Because many products and services that businesses and consumers
want are not controlled by a monopoly, most of the software industry long has
worked in an environment in which products from many vendors must interoperate.
That environment has engendered and sustained the enormous growth and
innovation in Internet computing and, before that, networked client-server
computing, which have fundamentally changed the way the commercial world does
business. Indeed, Microsoft has been nearly alone in doggedly resisting this
trend and insisting on proprietary protocols and interfaces that force users to
rely on a single vendor. Most of the interoperability loss on the Intel
platform today is due to Microsoft’s abuse of its dominant position to create
incompatibilities between its products and those of competitors.” Microsoft’s
goal is to own the computing platform and to create barriers to efficient
interoperation with its products for its competition,” IDC Microsoft on
Trial, supra, at 8, but that is not the inevitable course of the
software industry and contradicts the goals of federal antitrust law.
The idea that separating the Windows monopoly from the
Applications Company would impede innovation — as opposed to the mere
growth of Microsoft’s influence — is transparently absurd. See Jay Greene, Don’t
Worry, Bill — Innovation Will Survive, Bus. Wk., May 22, 2000, at 48.Unlike
the AT&T decree, plaintiffs’ proposal does not impose any
line-of-business restraints on either the Windows Company or the Applications
Company. Either company can develop any product that it wants, and the two can
compete with each other directly or indirectly. As University of California,
Berkeley professor and former Antitrust Division chief economist Richard
Gilbert has explained, by omitting the “innovation-stifling conditions that
were imposed on telecommunications,” Plaintiffs’ Proposed Final Judgment
“allows the marketplace, not the courts, to decide where the industry will go
next.” Richard Gilbert, A Better Breakup Than AT&T’s, N.Y. Times, May 10, 2000, at
A29.Almost all of the limited restrictions on the Windows Company’s conduct are
confined to business practices. The sole exception requires that any shackling
to Windows of middleware products (like those at the forefront of this case) be
accompanied by a simple add/remove utility, paralleled by a requirement of
economic neutrality to benefit those OEMs who choose to package the monopoly
product with competing middleware (or to leave that choice to the end-user).[1]See
Plaintiffs’ Proposed Final Judgment § 3(g). That preserves each company’s
ability to innovate, but avoids creating perverse incentives to intermingle
code in order to tie competitive products to monopoly products. See Lessig Br.
39 (Feb. 1, 2000). The burden on Microsoft is trivial, despite its
protestations, e.g., Microsoft Summ. Resp. 4-6, 48-52; Microsoft already
provides add/remove functionality for dozens of applications that are
“integrated” with Windows.
2. The
Proposed Reorganization Is A Moderate Approach To A Serious Structural Problem
Plaintiffs’ Proposed Final Judgment helps jump-start
competition with minimal disruption to Microsoft and minimal regulation of the
activities of the Windows Company and the Applications Company, while providing
Microsoft a significant ability to tailor the details for maximum efficiency
when it proposes its plan of reorganization under Section 1(a). The
reorganization proposed in Section 1(c) closely tracks Microsoft’s existing
organizational structure, which — despite Microsoft’s new, contrary claims
(Mem. Supp. Rejection 22-24) — differentiates platform products from
applications. See Greenhill & Williams Dec. ¶¶ 27-41; Microsoft
Reunifies Windows Divisions, Computer
Reseller News, Dec. 3, 1999,
http://www.techweb.com/wirestory/TWB19991203S0020.Microsoft’s protestations
about its “unitary” organization (Mem. Supp. Rejection 22) ring hollow in light
of its annual or more frequent voluntary restructuring, which make clear that a
restructuring along product lines would cause little incremental disruption.
See Greenhill & Williams Dec. ¶¶ 58-63; see Microsoft to Merge Two Main
Business Groups, Wall St. J.,
Mar. 31, 2000, at B6 (reporting that Microsoft will combine its operating
systems group with its development tools group). And the costs and dislocations
associated with the proposed reorganization are minor for a company of
Microsoft’s size and scope. See Greenhill & Williams Dec. ¶ 88.The Bell
System divestiture, which involved a million employees and billions of dollars
worth of complex and minutely integrated equipment spread across the entire
nation, was a far bigger undertaking than the reorganization of Microsoft.
Microsoft has about 35,000 employees, and its value is not in its physical
plant but in products that exist in computer code and supporting documentation.
Because Microsoft’s assets thus consist almost entirely of intellectual
property and personnel, both of which are easily transferred — and because the
principal tangible assets, computers, are easily moved — the proposed
reorganization would be substantially easier to accomplish than many other
voluntary or involuntary divestitures, including that of the Bell System. See
generally Economists’ Br. 58-60.
Microsoft is not the first huge monopolist to claim
that its company cannot possibly be divided to promote competition without
ruinous results. Ninety years ago, Standard Oil warned:
The inherent vice of this decree is that is seeks
to create an artificial division which never existed before; it does not seek
to compel members who were formerly independent to resume that independence,
but it seeks to compel different subcompanies, which have never been
independent, which have never been more than mere agencies created for
certain purposes, to sever their allegiance with the principal, and to stand
apart, independent and hostile to that principal and to each other.
Brief on the Law on Part of Appellants, Standard Oil Co. v. United
States, No. 725, Oct. Term, 1909, at 120-121 (emphasis added), quoted in William
E. Kovacic, Designing Antitrust Remedies for Dominant Firm Misconduct,
31 Conn. L. Rev. 1285,
1296-1297 (1999). Standard Oil likewise insisted that the units proposed for
divestiture all were
naturally a part of one whole — all operated together
and to and with each other — all were useful to the other, and to be so useful
must have a connection with one or more of the others. * * * There are many
parts, but each part has its place, and if a part is taken out, the whole
structure is disintegrated.
Id. at 284,
quoted in Kovacic, supra, at 1296.At oral argument, Standard reiterated,
“[A]ll these separate entities are parts of an organism, members of a single
great business. Tear them apart and who knows what will become of them * * * ?”
Oral Argument on Behalf of Appellants, Standard Oil Co. v. United
States, No. 398, Oct. Term, 1910, at 44, quoted in Kovacic, supra,
at 1297.Microsoft’s identical, overheated protests deserve no credence.
In fact, Plaintiffs’ Proposed Final Judgment reflects
a compromise between a full structural remedy that would directly create
competition or would flatly preclude Microsoft from competitively abusing
browser technologies, and one that would rely, futilely, on mere conduct
regulation alone. First, as the Economists’ Brief explains (at 48-52), the only
sure way to “terminate the illegal monopoly,” United Shoe, 391 U.S. at
250, and restore operating systems competition is to reorganize the Operating
Systems Business into three companies with equal assets competing head to head.
But Plaintiffs’ Proposed Final Judgment stops well short of that, choosing
instead to use milder means to “pry” the operating systems market “open to
competition,” Ford, 405 U.S. at 577-578, in a way that “enabl[es], but
[does] not compel[], competition to Windows.” Shapiro Dec.§ II (B), at
3.Second, although a requirement that the Windows Company obtain HTML rendering
only through third parties would be “easy to enforce” and “far from rocket
science,” Dan Gillmor, eJournal: News, Views and a Seattle Diary,
Mercury Center (San Jose Mercury News), Apr. 24, 2000,
http://weblog.mercurycenter.com/ejournal/2000/04/24 (quoting software developer
Dave Winer), Plaintiffs’ Proposed Final Judgment does not prohibit the Windows
Company from continuing to offer an Internet browser as a “Windows” component.
Plaintiffs’ proposal provides this additional leeway to the Microsoft
successors despite the competitive risks of allowing the Windows Company to
duplicate Microsoft’s behavior (so long as it uses any browser other than
Internet Explorer).
Thus, Plaintiffs’ Proposed Final Judgment surgically
removes an obstruction to competition, but relies on market forces to displace
Microsoft’s monopoly. Even Microsoft witness Gordon Eubanks told the New York
Times that this remedy would pose no danger to the industry and could produce
more competition on the desktop. See Steve Lohr & Joel Brinkley, Microsoft
Management Tells Workers There Will Be No Breakup, N.Y. Times, Apr. 26, 2000, at C1,
C9.That understates the remedy’s benefits, but correctly assesses the limited
risk of the reorganization.
The software market certainly has not shifted in a way
that makes a structural remedy less relevant. To the contrary, although the
demise of browser competition makes a conduct remedy directed at the browser
irrelevant, Microsoft’s success in its anticompetitive foray against Netscape
has strengthened Microsoft’s monopoly, see Henderson Dec. ¶ 41, has confirmed
Microsoft’s ability and incentive to misuse market power, and therefore has
made structural relief more necessary than ever. The multidirectionality of
Microsoft’s anticompetitive activity dooms specific conduct remedies to
irrelevance. Only by taking action to weaken Microsoft’s monopolistic hold can
a remedy hope to prevent recurrence of the type of anticompetitive conduct
proved in this case. At this moment, Microsoft is using its market power to try
to disadvantage Palm and RealNetworks to fend off those two potential threats
to the Windows monopoly. See p.22, supra.
Microsoft contends that a reorganization is
inappropriate because in no “contested case” has a court ordered a divestiture
from a “unitary” company. Mem. Supp. Rejection 4.[1]But
the reorganization here falls along established divisional lines that provide
an appropriate template for divestiture. See id. at 5.As part of its
annual restructuring, Microsoft last year placed all of its Windows development
groups in a single division. See, e.g., Microsoft Reunifies
Windows Divisions, supra. The only material alteration from this
design would be the transfer of Internet Explorer to the Applications Company,
the type of modest adjustment that is typical in antitrust remedies. If
Microsoft means to argue that divestiture is only appropriate when mergers have
been challenged, that obviously is not true. As the plaintiffs point out (Br.
31-32), the Supreme Court has directed or upheld divestitures when the exercise
of monopoly power, not its acquisition, was at issue. E.g., AT&T;
United Shoe, 391 U.S 244; Crescent Amusement, 323 U.S. at
188-190.And if Microsoft means that divestiture is not appropriate unless the
monopoly has been gained in part by acquisition, see Mem. Supp. Rejection
16-17, this case fits that description. Despite Microsoft’s claim that it has
not bolstered its position “by acquiring or merging with its rivals,” id.
at 17, Microsoft in fact acquired many of the components of Windows and Office,
and many other elements of its applications business, from other companies or
in the course of acquiring other firms.[1]In any
event, no monopolist, “unitary” or not, is — or should be — immune from the
only effective relief for monopolization. A structural monopoly problem
requires a structural solution no matter how the problem arose. As respected
commentators throughout the Nation have recognized, Plaintiffs’ Proposed Final
Judgment provides an appropriate solution for the violations in this case.[4]
3. Plaintiffs’
Proposed Final Judgment Will Not Delay Final Resolution Of This Case
Plaintiffs’ Proposed Final Judgment appropriately
calls for Microsoft to develop and propose a plan of reorganization to
implement the details of the divestiture decree. See Plaintiffs’ Proposed Final
Judgment § 1(a)-(c). In the plan, Microsoft can decide how to distribute
personnel in an orderly fashion, making the most efficient use of its current,
similar organizational structure. Its complaints about the lack of detail on
this score in Plaintiffs’ Proposed Final Judgment are misguided.
Entering Plaintiffs’ Proposed Final Judgment would
not, however, delay the speedy final resolution of this case. This Court need
not specify the details surrounding structural relief in order to render a
final judgment that may be certified for direct appeal to the Supreme Court
under the Expediting Act, 15 U.S.C. § 29(b). The Supreme Court has made clear
that an appeal — including an appeal under the Expediting Act — can proceed
once this Court enters its judgment identifying the relief, even though
resolution of the details of divestiture is deferred. Brown Shoe Co. v. United
States, 370 U.S. 294, 306-310 (1962); see Robert L. Stern et al., Supreme Court Practice 55 (7th ed. 1993). This is the
“settled course” in divestiture cases, in which the Supreme Court “has
consistently reviewed antitrust decrees contemplating either future divestiture
or comparable remedial action prior to the formulation and entry of the precise
details of the relief ordered.” See Brown Shoe, 370 U.S. at 309-310; see
also id. at 307 & n.14 (collecting cases).As in the AT&T
litigation, the plan of reorganization can proceed while the appeal is pending.
See, e.g., United States v. AT&T, 552 F. Supp. 131 (D.D.C.
Aug. 11, 1982) (decree entered), aff’d, Maryland v. United States,
460 U.S. 1001 (Feb. 28, 1983); United States v. Western Electric Co.,
569 F.Supp. 1057 (D.D.C. July 8, 1983) (plan of reorganization), aff’d, 464
U.S. 1013 (1983).
Plaintiffs’ Proposed Final Judgment appropriately
suggests (at § 6(a)) that the Court might stay implementation of the
divestiture but allow the conduct remedies to restrain Microsoft’s
anticompetitive acts while the appeal proceeds. As the Court has observed,
4/4/00 Tr. 11, 13, this is a clear case of “general public importance” that
warrants a direct appeal to the Supreme Court under the Expediting Act, 15
U.S.C. § 29(b). The legislative history of that provision confirms Congress’s
view that direct appeals are appropriate when an antitrust case has special
importance to the United States economy, regardless of the certiorari criteria
that apply in other cases. See, e.g., H.R. Rep. No. 93-1463, at 14
(1974); S. Rep. No. 93-298, at 4 (1974); see also Supreme Court Practice, supra, at 53-55.
This undeniably is such a case. Microsoft has
acknowledged “the lightening [sic] pace at which the software industry
moves” (Mem. Supp. Judgment 10), and has asserted both that software is “the
single most productive and envied industry in the United States” (Position as
to Proceedings 10) and that “Windows is very important to the Nation’s economy”
(Summ. Resp. 4). Microsoft also has claimed that, while “the threat of a
breakup hang[s] over [Microsoft’s] head,” Microsoft risks “los[ing]
irreplaceable employees” and “third parties may be unwilling to enter into
routine business agreements with Microsoft while its continued corporate
existence remains in doubt” (Mem. Supp. Rejection 5-6).
Microsoft thus is in no position to contend that the
software markets at issue here move slowly, that immediate resolution is
unnecessary, or that Microsoft’s domination of the desktop is not important to
the national economy. It is critical not only that the reorganization set forth
in Plaintiffs’ Proposed Final Judgment be implemented to restore competition to
the core of the software industry, but that competition supplant monopoly
before Microsoft grabs yet another market through its continuing campaign of
illegal conduct.[1]
Although Microsoft may legitimately seek to
cross-examine the plaintiffs’ witnesses on the remedy issue, that process
should not become a pretext for delaying this Court’s entry of a remedy and an
appealable final judgment. See Microsoft’s Position as to Proceedings, passim.[1]Most
important, this Court should not permit this remedy proceeding to become an
opportunity for Microsoft to harass and intimidate persons who have provided
the plaintiffs with information necessary to the investigation and conclusion
of this case. See id. at 8.Microsoft already is using the press to
spread word that it wants to use the discovery process to explore the thought
processes of enforcement personnel and the identities and communications of
witnesses. See Michael J. Martinez, Microsoft Seeks Federal Records, Denver Post, May 3, 2000, at C4; Joel
Brinkley & Steve Lohr, Microsoft to Dig In Against Breakup, New Orleans Times-Picayune, May 2,
2000, at C1. Microsoft has no legitimate claim on such discovery, however, even
if it were otherwise permissible. (It is not. See City of Burlington v. Westinghouse
Electric Corp., 246 F. Supp. 839, 847 (D.D.C. 1965) (denying similar
discovery to avoid fostering “a fear of exposure * * * which would hinder law
enforcement”); 5 U.S.C. § 552(b)(7) (excepting disclosure of
enforcement-related communications from Freedom of Information Act).)
Microsoft is thoroughly familiar with its own
corporate structure and its array of antitrust violations. It is untenable for
Microsoft to claim (Position as to Proceedings 8) that it had no inkling that
the plaintiffs might propose structural relief; the leading antitrust treatise
calls for such relief in almost all monopolization cases, 3 Areeda & Hovenkamp, supra, ¶
636c, at 56, and opinion leaders as staid as The Economist have been
calling for a breakup for months. See Now Bust Microsoft’s Trust, Economist, Nov. 13, 1999, at 15.Indeed,
the front page of Microsoft’s hometown newspaper reported the likelihood of a
breakup proposal long ago. See Joel Brinkley, The Goal: Break Up Monopoly;
‘Dramatic’ Microsoft Systems Remedy Sought, Seattle Post-Intelligencer, Nov. 10, 1999, at A1.Microsoft’s
transparent desire to use the judicial process to burden the plaintiffs’
sources of information should be seen for what it is: an effort to intimidate
witnesses and chill cooperation with the government.[1]It
is precisely because of improper efforts such as these that Plaintiffs’
Proposed Final Judgment (in Section 2(d)) specifically prohibits retaliation
against witnesses in this case. Sadly, Microsoft needs to be specifically
instructed not to commit a federal crime, see 18 U.S.C. §§ 1510-1513.
4. The
Benefits Of Restoring Competition Outweigh Any Risks
This Court should not be dissuaded from entering
Plaintiffs’ Proposed Final Judgment based on a concern about the destabilizing
effect of competition on a long-monopolized market. There is no reason to fear
that sort of change: to favor the stability of monopoly over the uncertainty of
competition would constitute a “frontal assault on the basic policy of the
Sherman Act.” Professional Engineers, 435 U.S. at 680.That policy
forecloses the claim that certain markets should be surrendered to
“monopolistic arrangements” rather than exposed to vigorous competition. Id.
at 690.
Consumers stand to gain the most from the increased
competition and innovation that would follow upon the full implementation of
Plaintiffs’ Proposed Final Judgment. See Romer Dec. ¶¶ 58, 61-62; Henderson ¶
28.Lower prices, greater reliability, and new functionalities all could be
expected to emerge on a far faster schedule than under a monopoly, where the
rationing of limited product improvements is critical to profit maximization.
The proposed reorganization not only would enhance competition and improve
innovation, as the Sherman Act is designed to do, but would do so in a way that
imposes no serious harm on shareholders or employees. Employees are empowered
by decentralized business structures with more room for advancement and
development. Microsoft’s bench is deep, and many talented engineers who have
been foreclosed from leadership roles would get the opportunity to use their
talents more fully.
Moreover, experience shows that divestitures often
increase shareholder wealth.[1]The
Standard Oil divestiture stands as a model remedy that broke up monopoly power,
created successor companies that eventually invaded each other’s original
territories and engaged in fierce competition, accelerated technological
advances, and enhanced shareholder value. See William S. Comanor & F.M.
Scherer, Rewriting History: the Early Sherman Act Monopolization Cases, 2 Int’l J. Econ. & Bus. 263,
266-268 (1995). Standard Oil, much like Microsoft today, had attacked
divestiture as “confiscatory” and “ruinous,” and said that the remedy “if
carried to its logical conclusions attacks the very foundations of the modern
business world.” Standard Oil Brief, supra, at 127, quoted in Kovacic, supra,
at 1298.Standard also warned that if the antitrust laws required monopolies to
break up, “prices will be higher; hundreds of thousands of men will be deprived
of employment; and our foreign trade * * * will be destroyed.” Standard Oil Argument,
supra, at 102-103, quoted in Kovacic, supra, at 1297.The
defenders of the Bell System monopoly attacked that divestiture in almost
identical terms. See Joel Brinkley, Microsoft Cites AT&T to Fight
Breakup, N.Y. Times. May
15, 2000, at C4 (noting similarities between Microsoft’s arguments and Bell
System contentions that monopoly was necessary for technological innovation and
that divestiture would primarily benefit Japanese competitors).
Barely a year after the Standard Oil divestiture,
however, Theodore Roosevelt remarked, “Wall Street’s prayer now is: ‘Oh,
Merciful Providence, give us another dissolution,’” Comanor & Scherer, supra,
at 266 (quoting Roosevelt), and the AT&T case produced similar
results. See, e.g., 100 Shares of the Old Ma Bell Today, Money, Aug. 1999, at 87 (100 shares of
AT&T, worth $6150 at divestiture, were worth $56,144 in mid-1999). In fact,
there are several striking resemblances between the Bell System’s position on
the eve of the divestiture decree in 1982 and Microsoft’s position now. The
Bell System considered itself “a single integrated enterprise” that was
“technologically integrated” and physically integrated nationwide. Brief of
AT&T, United States v. Western Electric Co., No. 87-5388, at
5 (D.C. Cir. filed July 25, 1989). See also Romer Dec. ¶ 14.That integration of
technologies and services into a monopolized (albeit regulated) base spurred a
government antitrust case that prompted a sweeping divestiture.
Although initially opposed by several States, the FCC,
the Department of Defense, and the Department of Commerce, and viewed with
trepidation by many consumers, the divestiture of the Bell System in fact
increased long distance and equipment competition dramatically. Interconnection
problems addressed by the case have proven to be readily manageable. The
telephone network is greatly improved and is a new center of commerce,
information, and entertainment, as well as ordinary conversation. The stunning
growth of the Internet and ubiquitous wireless telephony typify what happens
when competitive markets replace monopolized ones. And the divested fragments
of the old Bell System are all thriving, creating substantial wealth for
shareholders while facing new forces of competition in their particular
markets. See 100 Shares of the Old Ma Bell, supra.
It thus is not surprising that respected analysts have
recognized that a restructuring of Microsoft along the lines of the proposed
remedy, perhaps adding a second spin-off consisting of Internet Explorer and
other properties, may well maximize shareholder value.[1]Congress
properly determined that competition produces “the best allocation of our
economic resources, the lowest prices, the highest quality and the greatest
material progress.” Northern Pacific Ry. v. United States, 356
U.S. 1, 4 (1958). This Court should not hesitate to follow the course laid down
by the Supreme Court, and above all should not allow Microsoft to “game” the
judicial process by emerging from this case as it emerged from the consent
decree, confident that nothing material will change.
C. The
Browser Properties Should Be Separate
Plaintiffs’ Proposed Final Judgment goes far toward
providing an appropriate remedy for the problems of abused monopoly power that
have been worsened by the violations proved in this case. The Court should
modestly enhance that proposal, however, to deal directly with Microsoft’s
now-successful monopolization of the Internet browser market and its
concomitant threat to the competitive structure of Internet computing. Rather
than reorganizing Microsoft into only a Windows Company and a single
Applications Company, the Court should supplement Plaintiffs’ Proposed Final
Judgment by separating the Internet Explorer product and personnel into a
third, independent company. In the alternative, the Court should allow the
product and personnel to remain with the Applications Company, but should order
that the source code for the Internet Explorer product be disclosed and treated
as “open source” for the use of other software developers and vendors, ending
the monopoly position of the browser.
To allow a Microsoft successor to use the browser as a
point of leverage (or a means of protection) for its dominant applications
suite would reward Microsoft for its illegal conduct, even if the benefits did
not accrue to the operating systems business. Microsoft did not have a browser
monopoly when the trial began, but it acquired one in the meantime. Even during
a time period in which one would expect Microsoft to be most restrained, it
extended its market power.
1. If
Microsoft’s Applications Company retains Internet Explorer without limitation,
that Company will retain control over the critical link between the
productivity applications monopoly it now has and the monopolies over Internet
computing and the server market that Microsoft has sought to acquire. As
Microsoft executive Paul Maritz put it, “the most important thing we can do is
not lose control of the Web client,” because “[b]y controlling the client, you
also control the server.” Gov’t Ex. 498.The Applications Company could suppress
competition by adding features and protocols to the browser for which the
Applications Company has some particular need or advantage in implementation (e.g.,
through expertise or already existing software), and by refusing to add
features and protocols that would similarly advantage a competitor. Microsoft
expert Richard Schmalensee acknowledged this incentive at trial:“[I]f one
company controlled the browser and its look and feel and how it presented
applications, it could severely enhance or detract from the application
functionality of programs or applications running on the server.”6/24/99 (p.m.)
Tr. 46-47; see also id. at 48; Henderson Dec. ¶ 82 (quoting Rasmussen
Dep., 12/15/98 (a.m.), at 67-68).
Accordingly, rather than leave the monopoly browser
with the monopoly productivity applications, the remedy should separate the
three monopolies — operating systems, Internet browser, and desktop applications
— by spinning off an independent company to maintain and develop the Internet
Explorer product. That would remove the likelihood that the focus of the
desktop monopoly would simply shift from the operating system to the
productivity applications (with a welded-in browser). The Internet Explorer
Company would receive all intellectual property embodied in the current
stand-alone release of Internet Explorer that may be downloaded from
Microsoft’s website. The details of the transfer of the intellectual property
in Internet Explorer could parallel the provisions in Plaintiffs’ Proposed
Final Judgment (at § 1(c)(ii)), except that the intellectual property in
Internet Explorer would be conveyed to the Internet Explorer Company rather
than to the Applications Company (which in turn would receive a one-time
license like that afforded the Windows Company in Plaintiffs’ Proposed Final
Judgment § 1(c)(ii)). Likewise, and crucially, the Internet Explorer Company
would receive all Microsoft employees and contractors currently working on any
component of Internet Explorer, on the terms set out in Section 1(c)(i) of
Plaintiffs’ Proposed Final Judgment for the transfer of those employees and
contractors to the Applications Company.[1]
Once the Internet Explorer Company becomes
independent, software vendors would be on a more equal footing when it comes to
access to browser developments and compatibility between browsers and
applications. Given the importance of the browser as the universal client, the
universal interface to web-based (and other server-based) computing, see
Henderson Dec. ¶¶ 12, 81-86; Harris Direct ¶ 5, this parity is as critical as
the parity relating to the operating system. The safeguards applicable to the
Windows Company and the Applications Company, prescribed in Section 2 of
Plaintiffs’ Proposed Final Judgment, should also apply to the Internet Explorer
Company and to its relationship with the Windows Company and the Applications
Company. This supplement would have another benefit: the addition of a third
company with a key monopoly product will destabilize the relationship between
the Windows Company and the Applications Company, reducing the likelihood of
collusion (or the onset of tacit collusion) between the two.
The new Internet Explorer Company would have excellent
prospects. The Internet Explorer Company would instantly become one of the most
vigorous Internet companies, as the default supplier of the Internet user
interface to an installed base of more than 100 million Internet users. The
Internet Explorer Company surely would sponsor its own Internet portal, and
could sell placement on toolbars and channel bars. Potential content ventures
alone would be substantial for a holder of even a temporary 80-90% share of the
Internet browser market. The Internet Explorer Company could generate revenue
by providing engineering services to a wide variety of applications providers
that wanted to integrate their offerings with the leading browser. The Windows
Company and the Applications Company would be motivated to pay the Internet
Explorer Company to provide and maintain the functionality that they wanted to
invoke from within their products, at least until they developed competing
browsers of their own. The thriving applications service provider (ASP) market
would provide another source of revenue for the Internet Explorer Company. Asps
are particularly dependent upon the browser interface, and could be charged for
browser customization and for feature addition. OEMs and portals also might pay
the Internet Explorer Company to deliver customized browsers or to provide
browser integration with other offerings.
Even if the initial Internet Explorer Company ran into
difficulties, the Internet Explorer assets would not exit from the market. Many
software or content companies would jump at the chance to own the leading
browser software and market it in conjunction with their applications or
content businesses.
2. In
the alternative, if the Court desires to leave the Internet Explorer product
and personnel with the Applications Company as the plaintiffs propose, the
Court could limit the potential for anticompetitive use of the browser monopoly
by ordering that the Applications Company disclose the source code of the
Internet Explorer product and license the use of that code (and the code of any
successor products) on a full, “open source” basis. Although the Applications
Company ordinarily would be entitled to a reasonable royalty for intellectual
property subject to a compulsory licensing remedy, Microsoft has established
the reasonable commercial royalty for the Internet Explorer product at zero.
For that reason, and because the zero-royalty distribution of Internet Explorer
was an important part of Microsoft’s predatory campaign to foreclose competition
in the Internet browser market, it is appropriate to set the reasonable royalty
for Internet Explorer at zero for the purposes of an antitrust decree. If the
Court disagrees, the “reasonable” royalty under these circumstances should not
exceed a nominal amount, since the value of the product for Microsoft has
consisted in its anticompetitive effects.
The provision requiring open-source treatment of the
Internet Explorer product should be a continuing obligation to disclose the
source code in a “Timely Manner” within the meaning of Section 7(cc). The
obligation should have the duration specified in Section 3 of Plaintiffs’
Proposed Final Judgment, except that Microsoft and its successors should be
permanently enjoined from asserting intellectual property rights to any
Internet Explorer product source code that is disclosed pursuant to the decree.
3. Supplementing
Plaintiffs’ Proposed Final Judgment by adding an independent Internet Explorer Company
to the reorganization, or by making Internet Explorer an open source product,
would produce substantial incremental competitive benefits with little
additional disruption. In entering its decree, this Court therefore should make
that limited alteration.
D. Significant
Structural Relief In This Case Is Necessary To Preserve The Credible Deterrent
Force Of The Antitrust Laws
Given Microsoft’s demonstrated incentive and ability
to evade or to neutralize conduct restrictions, failure to require serious
structural relief would send a dangerous message to the business community and
the public in this highly publicized case. Conduct remedies alone would cost
Microsoft very little and would do little if any good for competition. The
practical message for other businesses with monopoly power and for the counsel
who advise them would be clear:
There are no structural remedies to fear. If Microsoft
was not subjected to structural relief, no one ever will be. A monopolist can
bank the advantages of its anticompetitive conduct, and deal with manageable
legal problems later when its goals have been achieved.
Unless there is a good-faith desire to comply with the
law, any company, including Microsoft, can and will skirt any conduct
prohibition that may be crafted.” Crown jewels” provisions are ineffective: if
structural relief is not sought and obtained on this record, it is hard to
imagine a court ordering a divestiture later based on the breach of a technical
decree obligation.
Without a meaningful divestiture remedy, Microsoft
will be free to leverage its operating system monopoly and extend it into
servers, Internet computing, and any other adjacent field that holds promise of
additional profit. No one will dare to bring antitrust violations to the
attention of the government if the results of this monumental suit are a
toothless “good conduct” code just like the 1995 consent decree. The public
deserves — and the law requires — an effective structural remedy in this
historic case.
The Court should enter
the Plaintiffs’ Proposed Final Judgment with the modification described above.
Respectfully submitted.
Edward J. Black, President (Bar No. 113282)
Ken Wasch, President (Bar No. 934984)
Jason Mahler (Bar No. 435605)
Software and Information Industry Association
Computer & Communications
1730 M Street, N.W.
Industry Association
Washington,
D.C. 20036
666 11th Street, N.W.
(202) 451-1600
Washington, D.C. 20001
(202) 783-0070
I hereby
certify that on this 19th day of May, 2000, I caused a true and correct copy of
the foregoing BRIEF ON REMEDY AS AMICI CURIAE OF THE COMPUTER AND
COMMUNICATIONS INDUSTRY ASSOCIATION AND THE SOFTWARE AND INFORMATION INDUSTRY
ASSOCIATION to be served by hand upon:
Antitrust Division
U.S. Department of Justice
950 Pennsylvania Avenue, N.W.
Washington, D.C. 20530
Richard J. Urowsky, Esq.
Sullivan & Cromwell
125 Broad Street
New York, NY10004
Fax: (212) 558-3588
Deputy Chief, Antitrust Bureau
New York State Attorney General’s Office
120 Broadway, Suite 2601
New York, NY 10271
Fax: (212) 416-6015
Microsoft Corporation
One Microsoft Way
Redmond, WA98052
Office of the Attorney General
of Wisconsin
123 West Washington Avenue
Madison, WI53703-7957
Christine Rosso, Esq.
Chief, Antitrust Bureau
Illinois Attorney General’s Office
100 West Randolph Street, 13th Floor
Chicago, IL60601
[2]The Best Is Yet to Come, remarks by Bill Gates to WINHEC 2000, New Orleans,
Apr. 25, 2000, http://www.microsoft.com/billgates/speeches/04?25winhec00.htm;
Henderson Dec. 43; see also, e.g., Gartner’s Dataquest Says Worldwide
PC Industry Experienced 15 Percent Growth in First Quarter 2000, press
release, Apr. 24, 2000,
http://gartner11.gartnerweb.com/dq/static/about/press/pr?b200019.html; Steve
Ballmer, address to Microsoft financial analyst meeting, July 22, 1999,
http://www.microsoft.com/msft/speech/analystmtg99/ballmerfam99.html.
[3]Although the AT&T remedy was the result of
a consent decree after the government and AT&T agreed to end the
litigation, the decision approving the decree is a persuasive precedent because
the remedy was thoroughly and vigorously disputed, and was analyzed in an
extensive opinion. Several States and the Department of Defense, among others,
opposed the remedy as excessive despite the agreement. Judge Greene treated the
case as fully contested, and relied on Supreme Court precedent laid down in
litigated cases. And the Supreme Court — which cannot and will not review cases
in which there is no live controversy — accepted the appeals of the States and
others and affirmed the decree on the merits. Maryland v. United
States, 460 U.S. 1001 (1983). See also California v. United
States, 464 U.S. 1013 (1983) (affirming plan of reorganization).
[4]See, e.g., Microsoft Corp., Microsoft
Enterprise Agreement, http://www.microsoft.com/CIO/licensing/agree.htm
(visited May 2000); Microsoft Corp. white paper, The Business Value of a
Microsoft Enterprise Agreement, July 1998,
http://www.microsoft.com/enterprise/licensing/agreement/Eawhite.doc; Steve
Ballmer, address to Microsoft financial analyst meeting, July 22, 1999,
http://www/microsoft.com/msft/speech/analystmtg99/ballmerfam99.htm (“the
consumer audience * * * is absolutely the smallest percentage of [Microsoft’s]
business”).
[4]See Microsoft Enterprise Agreement, supra;
Business Value, supra; Stuart Glascock & Barbara Darrow, Enterprise
Deals Sought — Microsoft, Resellers Broaden Scope of Enterprise Agreements, COMPUTER Reseller News, Apr. 20, 1998,
http://www.techweb.com/se/directlink.cgi?CRN19980420S0003.
[4]This brief does not attempt to identify and explain
specific improvements in the conduct provisions that might reduce the
possibility of evasion. The amici may seek leave to submit such
recommendations.
[4]Microsoft insists that the temporary conduct
provisions in Plaintiffs’ Proposed Final Judgment are too broad. See Summ. Resp.
23-55.If the Court finds any substance in Microsoft’s litany of complaints,
that would only confirm the inherent problems with formulating effective
conduct relief. Interpreting and enforcing the language of conduct provisions
will produce protracted litigation. Any effective antitrust remedy must rely on
structural requirements that cannot be litigated to a standstill.
[4]Of course, entirely different considerations apply to
requests for structural relief in private antitrust actions. See California
v. American Stores, Co., 495 U.S. 271, 295-296 (1990).
[4]Some commentators have raised concerns that the
reorganization could raise a “double marginalization” problem for consumers if
each monopolist raised its prices in a futile effort to appropriate a greater
share of the combined monopoly rents for Windows and Office — without taking
into account the impact of its price increase on the combined monopoly rents if
the other followed suit. We believe that the danger is overstated, largely for
the reasons expressed in the Shapiro Declaration (Section III(D), at 14-15).
Section 3(i) of Plaintiffs’ Proposed Final Judgment provides a safeguard for
three years after the reorganization by requiring Microsoft, upon any new
operating system release, to keep the immediately prior version of Windows
available at an unchanged price. If that temporary requirement also were
extended to Office (and to licensees other than OEMs), both Microsoft successor
companies would be competing sufficiently against themselves to restrain price
increases. If the Court remains troubled by the possibility of short-term price
increases before increased competition unfolds, however, the decree could cap
the prices of Windows and Office at their levels as of April 27, 2000, for the
first three years after implementation of the decree. In the alternative, the
decree could require the Windows Company and the Applications Company to award
each other one-time cross-licenses (at reasonable royalties) in Windows and
Office.
[4]The Court thus should disregard Microsoft’s hyperbolic
claim (Position as to Proceedings 1) that the disclosure provision amounts to
“wholesale confiscation of Microsoft’s intellectual property.” Section 3(b) of
Plaintiffs’ Proposed Final Judgment would not permit anyone to copy Microsoft’s
protected products, but simply would mandate the type of limited disclosure of
technical information that the Supreme Court endorsed in United States
v. National Lead, 332 U.S. at 353-360.The provision is carefully crafted
to do no more than prevent Microsoft from using the interfaces with its leading
monopoly product as a point of leverage into other markets or as a means of
punishing those who present or aid competitive challenges.
[4]The provision for a discount based on the proportion
of code devoted to the removed function further rewards Microsoft for efficient
application design. The less code needed for a middleware application, the
smaller the discount that an OEM would receive for removing it — reducing OEMs’
incentives to remove the Microsoft middleware in favor of a competing product.
[4]Of course, Microsoft would like to ignore cases like AT&T
that were litigated in part but that were settled by consent decree in light of
the expected results. This ignores the vigorous adversary debate, and the
lengthy opinions on the merits, that preceded Judge Greene’s order of
divestiture. See n.3, supra. See also, e.g., United States
v. United Fruit Co., 1958 Trade Cas. (CCH) ¶ 68,941, at 73,799 (Pt.
VIII) (D. La.) (consent decree, entered after years of litigation, required
defendant to divest, not an existing division, but a newly created company that
included assets, contracts, and personnel sufficient to create a sustainable
competitor to the defendant monopoly).
[5]The acquired products and technologies fall into four
major categories:
[5]Robert Bork, There’s No Choice: Dismember Microsoft, Wall. St. J., May 1, 2000, at A34;
Richard Gilbert, A Better Breakup Than AT&T’s, N.Y. Times, May 10, 2000, at A29; The
Remedy for Microsoft, N.Y.
Times, Apr. 28, 2000, at A22 (editorial); Bill Rockefeller?, Economist, Apr. 29, 2000, at 18
(editorial); Two Is Better Than One, Boston Globe, Apr. 28, 2000, at A22 (editorial); Unbroken
Windows, Christian Sci. Monitor,
May 1, 2000, at 10 (editorial); Microsoft, version 2.0: Structural Breakup
Is the Right Remedy, But How Many Pieces Would Be Enough?, S.J. Mercury News, May 1, 2000
(editorial); Dan Gillmor, Break Up Microsoft, S.J. Mercury News, Apr. 23, 2000,
http://www.mercurycenter.com/svtech/columns/gillmor/docs/dg042300.htm; Stephen
H. Wildstrom, A Win-Win-Win Breakup?, Bus. Wk., May 15, 2000, at 31.See also Does Microsoft
Stifle Innovation?, Bus. Wk.,
May 15, 2000, at 198 (editorial); Phillip Robinson, Microsoft Has Long
Carried A Big Stick to Bludgeon Innovators, Dallas Morning News, May 4, 2000, at 3F; Mike France, Fighting
for Market Share Is One Thing, Bus.
Wk., May 8, 2000, at 48.
[5]Thus, the Court should reject Microsoft’s request for
a toothless preliminary injunction that would permit a piecemeal appeal but
would further delay the final resolution of this case and the effective
remediation of Microsoft’s violations.
[5]Microsoft misrepresents (Position as to Proceedings
5-6) the scope of the remedy proceedings in National Lead — much less
any associated due process concerns. The large record to which the Supreme
Court referred was the entire trial record, one matched (if not
overmatched) in scope by the 78 days of testimony and thousands of exhibits in this
case. As the Court pointed out, a proposed “form of decree” was submitted after
the three-month trial in that case, 332 U.S. at 333-334; after the findings and
conclusions were entered, the additional proceedings on remedy consisted of
“further conferences with counsel.” Id. at 334.
[5]Microsoft’s announced plans already have succeeded in
chilling public comments by industry insiders who are concerned that Microsoft
will again seek depositions to support its contention that “the industry is
ganging up on” the monopolist. See Ephraim Schwartz, Proposed Microsoft
Breakup a Moot Point, Say Some Industry Observers, InfoWorld.com, Apr. 28,
2000,
http://www.infoworld.com/articles/pi/xml/00/04/28/000428pivendorreax.xml.And
Microsoft quickly moved to silence a software news Web site that exposed
Microsoft’s half-step palliation of the Kerberos problem that the plaintiffs’
remedy papers identified. See John Schwartz, Microsoft, Slashdot Exchange
Volleys, Wash. Post, May
12, 2000, at E1.
[5]If a structural remedy costs Microsoft or its
shareholders some money, those costs result from the necessary “den[ial] * * *
of the fruits” of the monopoly (United Shoe, 391 U.S. at 250) —
including the effects of past and anticipated monopoly rents on the stock
price. See Romer Dec. ¶¶ 69-70.“[T]he pinch on private interests is not
relevant to fashioning an antitrust decree.” Utah Public Service Comm’n
v. El Paso Natural Gas Co., 395 U.S. 464, 472 (1969) (disregarding
adverse tax effects of divestiture on defendant and shareholders). “[C]ourts are
authorized, indeed required, to decree relief effective to redress the
violations, whatever the adverse effect of such a decree on private interests.”
Du Pont, 366 U.S. at 326.Such effects serve goals of general deterrence
by motivating large shareholders and boards of directors to look more closely
at firms’ compliance with the antitrust laws — scrutiny woefully absent in this
instance.
[5]See, e.g., Mary Meeker (Morgan Stanley Dean
Witter), Judging Microsoft, Smartmoney.com, Apr. 5, 2000 (“Financially, we
believe a breakup could enhance shareholder value.”); Of Two Minds on
Microsoft, N.Y. Times,
Apr. 9, 2000, § 3, at 6 (quoting George Godfrey, ING Barings: “you’d win as a
shareholder”); Anthony Picardi & Dan Kusnetzsky, International Data Corp., Microsoft
on Trial: An Analysis of Possible Outcomes 7 (1999) (“[T]he long-term
growth of the company (and its progeny) and the health of the industry would be
best served if it were broken up”); Douglas Crook, Prudential Securities
analyst report, MSFT: Reiterate Strong Buy Rating and $145 Price Target,
Jan. 14, 2000 (“[W]e believe a divided Microsoft may have a higher market
capitalization than the combined entity.”).
[5]In submitting its plan of reorganization under
Plaintiffs’ Proposed Final Judgment § 1(a), Microsoft might propose to transfer
additional assets or employees to the Internet Explorer Company to maximize
innovation and shareholder value.