The Microsoft Antitrust Case*
by
Nicholas Economides**
Revised April 2, 2001
Abstract
This paper analyzes the law and economics of United States v. Microsoft, a landmark
case of antitrust intervention in network industries. The United States Department of
Justice and 19 States sued Microsoft alleging (i) that it monopolized the market for
operating systems of personal computers and took anti-competitive actions to illegally
maintain its monopoly; (ii) that it attempted to monopolize the market for Internet
browsers because such browsers would create competition for operating systems; (iii)
that it bundled its browser (Internet Explorer) with Windows; and that it engaged in a
number of other anti-competitive exclusionary arrangements with computer
manufacturers, Internet service providers, and content providers attempting to thwart the
distribution of Netscape’s browser. The District Court Judge found in most points for the
plaintiffs and ordered the breakup of Microsoft into two companies, one with all the
operating systems software, and one with all other products of the company. The District
Court also imposed a number of severe restrictions on the business conduct of Microsoft.
We analyze the economic issues related to liability. We also analyze the applicability
and effectiveness of the remedies imposed by the District Court and contrast them with
other potential remedies.
* Forthcoming in Journal of Industry, Competition and Trade: From Theory to Policy
(August 2001). I thank Gwill Allen, Gary Becker, Tim Brennan, Steve Davis, David
Evans, Frank Fisher, Rick Flyer, Eleanor Fox, Ron Gilson, Vic Goldberg, Eddie Lazear,
Roy Radner, Rick Rule, Dick Schmalensee, and participants at presentations at Canadian
Bureau of Competition, Columbia University, the Federal Reserve Bank of Philadelphia,
the Hastings School of Law, the New York University School of Law, Stanford
University (Economics), Stanford University (Graduate School of Business), the Stern
School of Business, the University of California (Berkeley), the University of California
(Los Angeles), and the University of California (Santa Barbara) for their comments and
suggestions. I am not a consultant of the United States Department of Justice, Microsoft,
or any of the Attorneys General of the 19 States and the District of Columbia that are
suing Microsoft.
** Stern School of Business, New York University, New York, NY 10012, and Stanford
University, Stanford, CA 94305, (212) 998-0864, fax (212) 995-4218,
http://www.stern.nyu.edu/networks/, neconomi@stern.nyu.edu
2
Contents
1. Introduction.............................................................................................................................................3
2. Markets With Network Effects ...............................................................................................................7
3. The Allegations.....................................................................................................................................10
4. Microsoft’s Defense..............................................................................................................................11
5. The Law................................................................................................................................................11
6. The Judge’s “Findings Of Fact” And “Conclusions Of Law” ..............................................................12
7. On Monopolization Of The Operating Systems Market .......................................................................13
a. Market Definition..............................................................................................................................13
b. Barriers To Entry ..............................................................................................................................14
c. Existence Of Market Power And Pricing Of The Operating System................................................15
d. Exercise Of Market Power................................................................................................................19
8. On Tying Of Windows With Internet Explorer ....................................................................................22
9. On Attempting To Monopolize The Browser Market...........................................................................24
10. Other Potentially Anti-competitive Arrangements ...........................................................................24
11. Effects On Consumers ......................................................................................................................24
12. Remedies...........................................................................................................................................27
a. Remedies As Imposed By The District Court...................................................................................27
b. Other Remedies Proposals ................................................................................................................31
13. Concluding Remarks.........................................................................................................................33
14. Bibliography .....................................................................................................................................35
3
The Microsoft Antitrust Case
1. Introduction
Microsoft is a large diversified computer software manufacturer. Microsoft
produces the Windows family of operating systems for personal computers and servers.
It also produces applications software that run on the Windows family of operating
systems, most notably the very successful MS-Office Suite consisting of Word (word
processor), Excel (spreadsheet), PowerPoint (presentations), Outlook (e-mail and news),
and Access (database).1
Almost all Microsoft products are complementary to a member
of the Windows family of operating systems for personal computers and servers.
During the last few years, the Federal Trade Commission and the Department of
Justice of the United States have investigated Microsoft on various antitrust allegations.
The 1991-1993 and 1993-1994 investigations by the Federal Trade Commission (“FTC”)
ended with no lawsuits. The 1994 investigation2
by the United States Department of
Justice (“DOJ”) was terminated with a consent decree in 1995.3
The key provisions of
the 1995 consent decree were:
1. Microsoft agreed to end “per-processor” (zero marginal price) contracts with
computer manufacturers (Original Equipment Manufacturers, “OEMs”) but it was
allowed to use unrestricted quantity discounts.
2. “Microsoft shall not enter into any License Agreement in which the terms of that
agreement are expressly or impliedly conditioned upon the licensing of any other
Covered Product, Operating System Software product or other product (provided,
however, that this provision in and of itself shall not be construed to prohibit
Microsoft from developing integrated products); or the OEM not licensing,
purchasing, using or distributing any non-Microsoft product.”4
Thus, the 1995 consent decree imposes two restrictions, one horizontal, and one
vertical. The horizontal restriction stops Microsoft from using zero marginal cost pricing.
1
Microsoft produces software, including operating systems for PC (Windows 95, 98, NT, 2000),
operating systems for local network and Internet servers (Windows NT, 2000), “back-office” products for
network and Internet servers, Internet clients, Internet and network servers, desktop applications (Office,
Word, Excel, Access, Outlook, PowerPoint, MS-Money, etc.), games, and programming languages (Visual
Basic, Java). Microsoft also produces services, including Internet service (MSN, WebTV), Internet content
(MSN), and product support, and some hardware such as branded mice, keyboards, etc.
2
DOJ sued Microsoft on July 15, 1994, under Section 2 of the Sherman Act, alleging that Microsoft
had entered into licensing agreements with OEMs that prevented other operating system vendors from
gaining widespread distribution of their products.
3
The Court entered the consent decree as its Final Judgment on April 21, 1995.
4
Final Judgment, Civil Action No. 94-1564.
4
However, it allows for quantity discounts, disregarding the fact that zero marginal cost
pricing is a special case of a quantity discount contract. 5
The vertical restriction of the 1995 consent decree prohibits product bundling
created by contract, but allows Microsoft to keep expanding the number and type of
functions of its products, including Windows. In short, in the 1995 consent decree
contractual bundling was disallowed, but technological bundling was explicitly allowed.6
During 1997, Senator Orin Hatch (R-Utah) held congressional hearings on
Microsoft that featured Microsoft’s CEO Bill Gates, Netscape’s CEO Jim Barksdale, Sun
CEO Scott McNealy, and PC manufacturer Michael Dell, among others. Senator Hatch
took the position that if present antitrust law cannot deal with various anti-competitive
acts attributed to Microsoft, Congress should change or enhance the antitrust laws.7
Sun
Microsystems, Oracle, IBM, Netscape, and Novell formed a loose coalition lobbying
intensely for antitrust action against Microsoft.8
On October 20, 1997, DOJ alleged that Microsoft violated the 1995 consent
decree by bundling Internet Explorer (“IE”) with the Windows operating systems, and
requiring computer manufacturers to distribute IE with Windows 95. DOJ petitioned the
District Court to find Microsoft in civil contempt. On December 11, 1997, Judge
Thomas Penfield Jackson issued a preliminary injunction barring the bundling of IE with
Windows.9
On May 12, 1998, the Court of Appeals (DC Circuit) voided the 1997
preliminary injunction. On June 23, 1998, the Court of Appeals ruled that the 1995
consent decree did not apply to Windows 98, which was shipped with an integrated IE as
5
This is a apparent contradiction in the terms of the consent decree. Clearly, the zero marginal
price contract is a special case of a quantity discount contract, more generally referred to by economists as
a non-linear pricing contract. If quantity discounts are allowed, but a zero marginal price is not (as the
consent decree reads), it is not clear how low the marginal price is allowed to be so that the consent decree
is not violated.
6
Microsoft has expanded over the years before and after the 1995 consent decree the functionality
included in Windows, leading to the elimination of some stand-alone add-ons markets. For example,
Microsoft included a disk defragmenter in Windows 1995 and the market for defragmenters promptly died.
Similarly, when hard disk compression was included in Windows 1995, the market for disk compression
software died. However, the market for fax software survived and expanded after the inclusion of fax
capabilities in Windows 1995.
7
Hatch (1999).
8
See, for example, Reback et al. (1994). Gary Reback represented Netscape and was instrumental
in creating this loose coalition, as well as providing the main arguments that the government used in this
case. Brinkley and Lohr (2000), page 326, also mention a Netscape “white paper” written by Wilson
Sonsini lawyers Gary Reback and Susan Creighton and entitled “White Paper Regarding the Recent
Anticompetitive Conduct of the Microsoft Corporation,” dated 1996, which was never made public but was
made available to the New York Times reporters.
9
The Court also referred the issue to a special master, Prof. Lawrence Lessig of Harvard.
5
part of the operating system and an IE icon on the PC desktop, arguing that “courts are ill
equipped to evaluate the benefits of high-tech product design.”10
During the week following the Court of Appeals stay of Judge Jackson’s
preliminary injunction that barred the bundling of IE with Windows because of the
alleged violation of the 1995 consent decree, DOJ filed a major antitrust suit against
Microsoft. In this action (DOJ Complaint 98-12320), filed on May 18, 1998, DOJ was
joined by the Attorneys General of 20 States and the District of Columbia. This paper
focuses on this last and continuing lawsuit against Microsoft.
Over the years, Microsoft has integrated in the Windows class of operating
systems many functions and features that were originally performed by stand-alone
products. 11
Moreover, the Court of Appeals in its June 23, 1998 decision affirmed that
Microsoft’s practice of bundling IE with Windows was legal under the terms of the 1995
consent decree. To overcome this interpretation of the law, DOJ argued that Microsoft’s
bundling of IE with Windows and its attempt to eliminate Netscape as a competitor in the
browser market was much more than adding functionality to Windows and marginalizing
a series of add-on software manufacturers. DOJ alleged (and the District Court
concurred) that Microsoft added browser functionality to Windows and marginalized
Netscape because Netscape posed a potential competitive threat to the Windows
operating system. This distinctive threat posed by Netscape was a crucial part of the DOJ
allegations. DOJ alleged that applications could be written to be executed “on top” of
Netscape; since Netscape could be run on a number of operating systems, DOJ alleged
that Netscape could erode the market power of Windows. In DOJ’s logic, Microsoft gave
away IE and integrated it in Windows so that Netscape would not become a platform that
would compete with Windows. Thus, DOJ alleged that Microsoft’s free distribution of
IE, its bundling with Windows, and all its attempts to win the browser wars were
defensive moves by Microsoft to protect its Windows monopoly.
The Microsoft trial took place at an accelerated schedule at the U.S. District Court
of the District of Columbia from October 19, 1998 to June 24, 1999. Only twelve
witnesses testified from each side. Microsoft’s CEO Bill Gates was not called as a
witness, but his video taped deposition was extensively used during the trial. Judge
Jackson pre-announced that he would announce his “findings of fact” before his
“conclusions of law.” This was widely interpreted as implying that the judge was trying
to give an opportunity to the sides to reach a compromise and resolve the case through a
consent decree.
On November 5, 1999, Judge Jackson issued his “findings of fact,” siding very
strongly with the plaintiffs. In December 1999, Judge Richard Posner, a prominent
10
The Court of Appeals further noted that “the limited competence of courts to evaluate high-tech
product designs and the high cost of error should make them wary of second-guessing the claimed benefits
of a particular design decision.” 147 F.3d at 950 n.13.
11
For example, disk compression and disk de-fragmentation were not part of Windows 3.1 and were
added to Windows 95 and 98.
6
antitrust scholar and the Chief Judge of the Seventh Circuit Court of Appeals, agreed to
serve as mediator for settlement discussions.12
On April 1, 2000, settlement talks broke
down after some States reportedly disagreed with the proposed agreement.13
On April 3,
2000, Judge Jackson issued his “conclusions of law” finding for the plaintiffs on almost
all points. In particular, Judge Jackson found Microsoft liable for monopolization and
anti-competitive tying of IE with Windows but found that Microsoft’s exclusive contracts
did not make it liable for preventing Netscape from being distributed.
On June 7, 2000, after an extremely short hearing,14
Judge Jackson issued his
remedies decision, splitting Microsoft into two companies, and imposing severe business
conduct restrictions. Microsoft appealed, and was granted a stay of all parts of the
District Court decisions until the appeal is heard. Although the Washington DC Appeals
Court expressed its willingness to hear the case in plenary session, the District Court
agreed with the government’s proposal to petition the Supreme Court to hear the case
immediately, invoking a rarely used provision of antitrust law.15
On September 26,
2000, the Supreme Court indicated that it will not hear the case before the Court of
Appeals. Thus, the appeals process is likely to take at least a year, and more likely two
years. It is widely believed that the Federal lawsuit will be settled during the
administration of president George W. Bush, who had stated during his election
campaign that he will only pursue price fixing antitrust cases.16
The nature and scope of the allegations in this case imply that its final outcome
will determine the terms of competition in the software industry and more generally in
network industries. To a large extent, the application of antitrust law in network
industries is hampered by the fact that rigorous economic analysis of network industries
is relatively new.
The rest of this paper has the following structure. Section 2 discusses general
market structure issues in industries with network effects. Section 3 outlines the
allegations in the Microsoft case. Section 4 outlines Microsoft’s defense. Section 5
summarizes the key sections of antitrust law relevant to this case. Section 6 outlines the
District Court Judge’s “findings of fact” and “conclusions of law.” Section 7 analyzes
the allegation of monopolization of the operating systems market. Section 8 analyzes the
12
As mediator, Judge Posner was not acting in his judicial capacity.
13
See New York Times, April 2, 2000.
14
The hearing was on May 24, 2000. It started at 10:15 am, ended around 3:30 pm, and included a
two-hour lunch break.
15
DOJ argued that review by the Supreme Court was appropriate to expedite the final judgment
because of the importance of the case for the national economy. However, it is worth noting that the
plaintiffs had a real interest to avoid the DC Court of Appeals since that Court ruled (June 23, 1998) on the
issue of bundling Internet Explorer and Windows 95 in the earlier Microsoft case exactly in the opposite
way to Judge Jackson’s decision on bundling in the case being appealed.
16
Financial Times, June 26, 2000, Interview with George W. Bush.
7
allegation of tying Windows with IE. Section 9 analyzes the allegation of attempted
monopolization of the browser market. Section 10 discusses other potentially anticompetitive
arrangements. Section 11 analyzes the effects on consumers. Section 12
discusses remedies. Section 13 has concluding remarks.
2. Markets With Network Effects
In assessing the Microsoft case, it is important to remember that the case focuses
on markets with network effects. Network effects define crucial features of market
structure that have to be taken into consideration in understanding competition and
potentially anti-competitive actions in these markets.
A market exhibits network effects (or network externalities)17
when the value to a
buyer of an extra unit is higher when more units are sold, everything else being equal. In
a traditional network, network externalities arise because a typical subscriber can reach
more subscribers in a larger network.18
In a virtual network,19
network externalities arise
because larger sales of component A induce larger availability of complementary
components B1, ..., Bn, thereby increasing the value of component A. The increased
value of component A results in further positive feedback.20
For example, the existence
of an abundance of Windows-compatible applications increases the value of Windows.
There are a number of crucial features of markets with network effects that
distinguish them from other markets. First, markets with strong network effects where
firms can chose their own technical standards are “winner-take-most” markets. That is,
in these markets, there is extreme market share and profits inequality.21
The market share
of the largest firm can easily be a multiple of the market share of the second largest, the
second largest firm’s market share can be a multiple of the market share of the third, and
so on. This geometric sequence of market shares implies that, even for small n, the nth
firm’s market share is tiny.
For example, abundance of applications written for Windows increases the value
of Windows and induces more consumers to buy Windows. This increases the incentive
17
The word externality means that a good’s value is not intermediated in a market. For the purposes
of this paper, we will use the words “network effects” and “network externalities” interchangeably.
18
See Economides (1996).
19
A virtual network is a collection of compatible goods (that share a common technical platform).
For example, all VHS video players make up a virtual network. Similarly, all computers running Windows
98 can be thought of as a virtual network.
20
Despite the cycle of positive feedbacks, it is typically expected that the value of component A
does not explode to infinity because the additional positive feedback is expected to decrease with increases
in the size of the network.
21
See Economides and Flyer (1998).
8
for independent applications writers to write applications for Windows, and this further
increases sales and market share for Windows. Moreover, consumers are willing to pay
more for the brand with the highest market share (since it has more associated
applications), and therefore profits associated with this brand can be a large multiple of
profits of other platforms. This implies a very large market share for Windows, a small
market share for the Mac, a very small market share for the third competitor, and almost
negligible shares for the fourth and other competitors.
Second, due to the natural extreme inequality in market shares and profits in such
markets at any point in time, there should be no presumption that there were anticompetitive
actions that were responsible for the creation of the market share inequality
or the very high profitability of a top firm. Great inequality in sales and profits is the
natural equilibrium in markets with network externalities and incompatible technical
standards. No anti-competitive acts are necessary to create this inequality.22
Third, because “winner takes most” is the natural equilibrium in these markets,
attempting to superimpose a different market structure, (say one of all firms having
approximately equal market shares), is futile and counterproductive. If a different market
structure is imposed by a singular structural act (say a breakup of a dominant firm), the
market would naturally deviate from it and instead converge to the natural inequality
equilibrium. If forced equality is imposed as a permanent condition, it would create
significant social inefficiency, as discussed below.
Fourth, in network markets, once few firms are in operation, the addition of new
competitors, say under conditions of free entry, does not change the market structure in
any significant way. The addition of a fourth competitor to a triopoly hardly changes the
market shares, prices, and profits of the three top competitors.23
This is true under
22
Litan et al. (2000) err in reasoning that Microsoft’s very high profitability is a clear indication of
monopolization in the antitrust sense. High profitability for the top platform is natural in this winner-takemost
market.
23
See Economides and Flyer (1998). The table below, taken from this paper, shows market
coverage and prices as the number of firms with incompatible platforms increases. Maximum potential
sales was normalized to 1.
Table 1: Quantities, Market Coverage, And Prices Among Incompatible Platforms
Number
of firms
I
Sales of
largest
firm
q1
Sales of
second
firm
q2
Sales of
third
firm
q3
Market
coverage
ΣI
j=i qj
Price
of
largest firm
p1
Price of
second
firm
p2
Price of
third
firm
p3
Price
of smallest
firm
pI
1 0.6666 0.6666 0.222222 2.222e-1
2 0.6357 0.2428 0.8785 0.172604 0.0294 2.948e-2
3 0.6340 0.2326 0.0888 0.9555 0.170007 0.0231 0.0035 3.508e-3
4 0.6339 0.2320 0.0851 0.9837 0.169881 0.0227 0.0030 4.533e-4
5 0.6339 0.2320 0.0849 0.9940 0.169873 0.0227 0.0030 7.086e-5
6 0.6339 0.2320 0.0849 0.9999 0.169873 0.0227 0.0030 9.88e-11
7 0.6339 0.2320 0.0849 0.9999 0.169873 0.0227 0.0030 0
9
conditions of free entry. Therefore, although eliminating barriers to entry can encourage
competition, the resulting competition does not significantly affect market structure. In
markets with strong network effects, antitrust authorities cannot significantly affect
equilibrium market structure by eliminating barriers to entry.
Fifth, the fact that the natural equilibrium in network industries is winner-takemost
with very significant market inequality does not imply that competition is weak.
Competition on which firm will create the top platform and reap most of the benefits is,
in fact, very intense.
Sixth, there is a more fundamental concern about the application of antitrust in
network industries.24
In industries with significant network externalities, under
conditions of incompatibility between competing platforms, monopoly may maximize
social surplus. When strong network effects are present, a very large market share of one
platform creates significant network benefits for this platform which contribute to large
consumers’ and producers’ surpluses. It is possible to have situations where a breakup of
a monopoly into two competing firms of incompatible standards reduces rather than
increases social surplus because network externalities benefits are reduced. This is
another way of saying that de facto standardization is valuable, even if done by a
monopolist.25
Seventh, in network industries, the costs of entry may be higher but the rewards of
success may also be higher compared to non-network industries. Thus, it is unclear if
there is going to be less entry in network industries compared to traditional industries. If
a requirement for entry is innovation, one can read the previous statement as saying that it
is unclear if innovation would be more or less intense in network industries. The
dynamics of the innovation process in the winner-take-most environment of network
industries are not sufficiently understood by academic economists so that they could give
credible advice on this issue to antitrust authorities. However, in the last two decades we
Note that the addition of the fourth firm onward makes practically no difference in the sales and prices of
the top three firms.
24
In the Microsoft case, both sides had the chance to address this issue, but failed to do so.
25
Economides and Flyer (1998), show that, in market conditions similar to the ones in the OS
software market, social welfare (total social surplus) can be higher in monopoly. The table below, taken
from this paper, shows profits, consumers’ and total surplus in a market where firms produce incompatible
products, as the number of competitors I increases.
Table 2: Profits, Consumers’ And Total Surplus Among Incompatible Platforms
Total
number
of firms I
Profits of
largest firm
Π1
Profits of
second firm
Π2
Profits of
third firm
Π3
Total industry
profits
ΣI
j=1 Πj
Consumers’
surplus
CS
Total
surplus
TS
1 0.1481 0.1481 0.148197 0.29629651
2 0.1097 7.159e-3 0.1168 0.173219 0.29001881
3 0.1077 5.377e-3 3.508e-4 0.1135 0.175288 0.28878819
10
have observed very intense competition in innovative activities in network industries
financed by capital markets.
Eighth, the existence of an installed base of consumers favors an incumbent.
However, competitors with significant product advantages or a better pricing strategy can
overcome the advantage of an installed base.26
Network effects intensify competition,
and an entrant with a significantly better product can unseat the incumbent. In network
industries, we often observe Schumpeterian races for market dominance. This is a
consequence of the winner-take-most natural equilibrium combined with the high
intensity of competition that network externalities imply.
3. The Allegations
On May 18, 1998, the United States Department of Justice, and the Attorneys
General of 20 States27
and the District of Columbia sued Microsoft. The main allegations
were:
1. Microsoft illegally monopolized the market for operating systems (“OSs”) for
personal computers (“PCs”) under ¶ 2 of the Sherman Antitrust Act;
2. Microsoft had anti-competitive contractual arrangements with various vendors of
related goods, such as with computer manufacturers (“OEMs”) and Internet
Service Providers (“ISPs”), and had taken other actions to preserve and enhance
its monopoly; that these contractual arrangements and other actions were illegal
under ¶ 2 of the Sherman Antitrust Act;
3. Microsoft illegally attempted to monopolize the market for Internet browsers (but
failed to succeed), an act that is illegal under ¶ 2 of the Sherman Antitrust Act;
4. Microsoft bundled anti-competitively its Internet browser, IE, the Microsoft
Internet browser, with its Windows operating systems; that this is illegal under ¶ 1
of the Sherman Antitrust Act.
26
A clear example of this is the win of VHS over Beta in the United States consumer video
recorders market. Beta was first to market and had a significant installed base in the five years of the
coexistence of the two competing standards. However, because VHS (i) introduced earlier a recording tape
of longer duration; (ii) used wide and inexpensive licensing of its technology; and (iii) its licensees had a
much wider distribution system, VHS emerged as the winner, and Sony stopped selling Beta recorders to
the US consumer market.
27
The plaintiff States were California, Connecticut, Florida, Illinois, Iowa, Kansas, Kentucky,
Louisiana, Maryland, Massachusetts, Michigan, Minnesota, New Mexico, New York, North Carolina,
Ohio, South Carolina, Utah, West Virginia and Wisconsin. Originally South Carolina was part of the
lawsuit but dropped out of the case soon after the acquisition of Netscape by America Online (“AOL”) and
Sun Microsystems (“Sun”) stating that this acquisition has eliminated any anti-competitive concerns.
11
4. Microsoft’s Defense
Microsoft’s defense was as follows. First, Microsoft argued that the law was on
its side since the Court of Appeals had ruled on June 23, 1998 that Microsoft can legally
add new features and functions to Windows. Therefore Microsoft argued that it was legal
to add IE’s functionality to Windows, and it had done nothing wrong by integrating IE in
Windows.
Second, Microsoft argued that it was just competing hard against Netscape, that
such competition was welfare-enhancing, and that it did not commit any anti-competitive
acts.
Third, Microsoft argued that it did not have monopoly power in the operating
systems market.
Fourth, Microsoft argued that competition in the software sector was intense and
that its leadership position could be replaced at any time by a new competitor or entrant.
Although this seems to have been a deeply held belief of Microsoft’s management (as
revealed by the internal e-mails), its economics expert witness failed to convincingly
articulate this Schumpeterian view in the context of an antitrust defense.
Fifth, Microsoft argued that it is a leader in software innovation and that it has
enhanced rather than hobbled the innovation process.
Sixth, Microsoft argued that consumers have benefited from its actions rather than
been harmed by them. Microsoft claimed direct consumer benefits from its low pricing
of the operating system, the zero pricing of its Internet browser, and from its
enhancement and acceleration of the innovation process. Microsoft also argued (rather
ineffectively) that consumers benefit from the de facto standardization that its large
market share brought to the operating systems market.
5. The Law
The analysis of DOJ’s allegations and Microsoft’s defenses requires an
examination of antitrust law as it currently applies in the United States, the facts of the
case, and a synthesis that will define the extent of the violations of the law. We start with
the examination of antitrust law as it applies to this case.
We review the antitrust law in the areas that apply to the Microsoft case. Section
2 of the Sherman Antitrust Act states:
“Every person who shall monopolize, or attempt to monopolize, or combine or
conspire with any other person or persons, to monopolize any part of the trade or
commerce among the several States, or with foreign nations, shall be deemed
guilty of a felony.”
12
The U.S. antitrust law, as presently interpreted, implies that “monopolization”
under ¶ 2 of the Sherman Act is illegal if the offender took anti-competitive actions to
acquire, preserve, or enhance its monopoly. For “monopolization,” plaintiffs have to
prove that the defendant
1. Possessed market power; and
2. willfully acquired or maintained this monopoly power as distinguished from
acquisition through a superior product, business acumen, or historical accident.
Therefore, contrary to popular belief, for monopolization to be illegal under U.S. antitrust
law, it is not sufficient for a company to “monopolize” a market in the sense of
possessing a very large market share, even a market share of 100%.
U.S. antitrust law, as presently interpreted, implies that “attempting to
monopolize” is illegal under ¶ 2 of Sherman Antitrust Act if the specific actions taken
have anti-competitive consequences. For example, bundling, and, more generally, price
discrimination could be illegal if they have anti-competitive consequences. Similarly,
exclusionary contracts (which restrict distribution or production) could be illegal if they
have anti-competitive effects.
To prove “attempting to monopolize” (under Sherman Act ¶ 2), plaintiffs have to
prove that the defendant
1. Engaged in predatory or anti-competitive conduct
2. with specific intent to monopolize
3. and that there was a “dangerous probability” that the defendant would succeed in
achieving monopoly power.
Section 1 of the Sherman Antitrust Act states:
“Every contract, combination in the form of trust or otherwise, or conspiracy, in
restraint of trade or commerce among the several States, or with foreign nations,
is declared to be illegal.”
Thus, unreasonable “restraint of competition” is illegal under ¶ 1 of the Sherman Act; this
may include tying of products or other exclusive arrangements.
It has generally been accepted in the application of antitrust law in the United
States that anticompetitive acts harm consumers. Consumer harm or attempted harm is
necessary for an antitrust violation. The affected consumer group may be present or
future. Consumers may lose directly from high prices or indirectly through a limitation
of choices of variety and quality or by a retardation of the innovation process. But
without consumer victims or prospective likely victims, present or future, it is extremely
hard to prove that an antitrust violation exists.
6. The Judge’s “Findings Of Fact” And “Conclusions Of Law”
13
The U.S. District Court’s “findings of fact” (November 1999) and “conclusions of
law” (April 2000) find for the plaintiffs (U.S. Department of Justice and 19 States) in
almost all the allegations against Microsoft. In particular, Judge Penfield Jackson found:
1. The relevant antitrust market is the PC operating systems market for Intelcompatible
computers.
2. Microsoft has a monopoly in this market “where it enjoys a large and stable
market share.”
3. Microsoft’s monopoly is protected by the “applications barrier to entry,” which
the judge defines as the availability of an abundance of applications running
Windows.
4. Microsoft used its monopoly power in the PC operating systems market to
exclude rivals and harm competitors.
5. Microsoft hobbled the innovation process.
6. Microsoft’s actions harmed consumers.
7. Various Microsoft contracts had anti-competitive implications, but Microsoft is
not guilty of anti-competitive exclusive dealing contracts hindering the
distribution of Netscape Navigator.
7. On Monopolization Of The Operating Systems Market
a. Market Definition
The definition of a market for antitrust purposes is crucial for the determination of
liability. Traditionally, an antitrust market definition is based on substitution
considerations. To define the market one has to examine the product. Windows was sold
to OEMs and users as an operating system on which they could run software applications.
Windows also contained a variety of functions that were provided free of charge to
independent software vendors (“ISVs”) so that they would find it easier, cheaper, and
faster to create applications for Windows. Microsoft sold to ISVs various tools so that
they would utilize these functions of Windows, and provided a variety of free services to
ISVs. Clearly ISVs were subsidized in kind so that they would write applications for
Windows. DOJ and the District Court decided to ignore the market relationship between
ISVs and Microsoft and to focus exclusively on the sale of Windows to OEMs.
DOJ proposed and the District Court Judge agreed that the relevant antitrust
market was the market for operating systems for personal computers that are based on an
Intel-compatible Central Processing Unit (“CPU”). In this respect, the judge took a
rather narrow and static view of the market. It is clear that consumers demand
applications services from personal computers and would be willing to switch their
purchases to non-Intel-compatible PCs if these computers were offered with similar
applications as Intel-based PCs and at attractive prices. Thus, non-Intel-compatible
computers, operating systems and software are substitutes for Intel-compatible
computers, operating systems and software. If one takes a static view and considers the
software applications written for each computer and operating system as given and
unchanging (as the judge did), then there is no close substitute for Windows that can run
14
as many applications as Windows does. In this static view, all substitutes for Windows
are much less valuable because they can run so few applications compared to Windows.
In a more dynamic view, if new software applications can be written for an operating
system (or old ones originally written for other operating systems “ported to” it), an
operating system can easily have close substitutes.
The dynamic view of market definition invites an examination of the incentives
that ISVs have to create such applications. Their incentives are not just defined by the
price of the operating system to OEMs and final consumers, and by the number of units
that the operating system is expected to sell over the lifetime of their applications. ISVs
use functions that are embedded in an operating system or platform to perform certain
operations that are necessary for their applications. The more such functions are
available in an operating system, and the easier they are to access, the more likely it is for
an ISV to write an application for the OS. Thus, the availability, abundance, and ease of
use of embedded functions in the operating system that can be utilized by an ISV needs to
be calibrated in order to define the extent of the market in the dynamic view of market
definition.
Early in the trial, Microsoft argued that the concept of market definition for
antitrust purposes was inappropriate for the dynamic setting of software markets. This
position was not very well articulated, and hardly appropriate for a District Court Judge
who was very unlikely to subscribe to a major revision to traditional application of
antitrust law. Later in the trial, Microsoft argued through its economic witness that
platforms rather than operating systems were the appropriate domain for antitrust
analysis, stressing the fact that functions embedded in platforms are used by ISVs to cut
their costs of writing applications. However, a full and coherent view of a dynamic
market definition was never presented.
b. Barriers To Entry
A crucial tenet of DOJ’s and the District Court’s argument on monopolization
was that there are significant barriers to entry in the market for operating systems for
personal computers. Barriers to entry are asymmetric costs that entrants incur but are not
incurred by incumbents. Software has a very large fixed cost and a very low (almost
zero) marginal cost. The fixed cost required to create an operating system could create a
barrier to entry to the extent that it is sunk by the incumbent(s), while it has to be paid by
an entrant in the current period. However, this barrier to entry is not very significant,
given the size of the market.
The main barrier to entry could be created by the abundance of applications that
run on Windows. It is estimated that Windows runs 70,000 applications, while the
Macintosh and the Linux operating systems run far fewer applications. Currently, there
are also applications that run on servers connected with the World Wide Web and are
essentially independent of the operating system (Windows, Mac, Linux, etc.) of the
client. The number of applications that can be run on an operating system can be thought
of as a quality index. What determines the existence and the extent of the “applications
15
barrier to entry” is the extent that costs incurred today by entrants are already sunk by the
incumbent. The fact that an industry has a high fixed cost or requires investment to
induce others to create complementary goods, by itself is not enough to conclude that
there are barriers to entry. To judge whether barriers to entry exist, one has to examine if,
in the relevant time frame, the entrants are facing asymmetrically higher costs than the
incumbent(s).
It is clear that, early in its history, Microsoft understood the importance of an
abundance of applications for its operating systems. To achieve this, Microsoft provided
various inducements to independent applications developers to write applications for its
operating systems. A key inducement for independent applications developers has been
the availability of functions in Windows that are useful for applications developers. For
example, Windows and other operating systems include programming code that makes it
easy for an application to print to any of a large number of printers.28
Since this function
is available in the operating system, applications developers do not have to write their
own programming code to accomplish this function. It follows that applications are now
cheaper and quicker to write.
Microsoft spends considerable resources to attract independent software
developers to write for its operating system and to include in the operating system a
variety of functions that they find useful. It is clear that the abundance of applications for
Windows makes Windows more valuable, and therefore increases Microsoft’s potential
profits. This is clearly a pro-competitive reason for Microsoft to invest in resources to
attract independent software developers. Given the same number of competitors in the
operating systems market, a firm enhances its competitive position by investing resources
in order to attract applications to its operating system. At the same time, such an increase
in the number of applications available for an operating system makes it very difficult for
new entrants to profitably enter and survive. Thus, increasing the number of applications
available for an operating system, although it may have a purely competitive justification,
both increases the cost of entry in the operating systems market, and creates the potential
for anti-competitive exploitation.
On the other hand, to the extent that the incumbent is currently spending
resources to attract applications to its operating system, the difference between the costs
faced by the incumbent and a potential entrant is reduced, and therefore barriers to entry
are reduced. Thus, the existence, extent, and importance of an “applications barrier to
entry” depends on an empirical examination of the extent to which Microsoft spends
resources in the current time frame to attract applications to Windows, and the extent to
which Microsoft has already sunk such costs in past time periods. Such an examination
was not performed by the District Court.
c. Existence Of Market Power And Pricing Of The Operating System
28
This function was not available in DOS. Thus, each application written for DOS had to include
large numbers printer drivers that made sure that the application would be able to print on users’ printers.
16
The judge ruled that Microsoft has monopoly power in the OS market for Intelcompatible
PCs. In antitrust, it is generally understood that a firm has monopoly power
when it has the sustained ability to control price or exclude competitors. The existence of
significant barriers to entry and the very high market share of Microsoft in the operating
systems market gave indications that Microsoft had monopoly power. But there was also
a very strong indication to the contrary. Microsoft priced its operating system to Original
Equipment Manufacturers (“OEMs”) at an average price of $40-60, a ridiculously low
price compared to the static monopoly price.29
Microsoft’s economic witness showed
that the static monopoly price was about $1,800, a large multiple of Microsoft’s actual
price.30
At first glance, it seems that Microsoft could not possibly have monopoly power
in OSs when its OS price is about 3% of the monopoly price.
Understanding and explaining the very low price of Windows is important for
understanding what Microsoft’s competitive position was and how Microsoft thought of
it. Plaintiffs failed to explain Microsoft’s pricing by either assuming it away or giving
spurious explanations. First, the government claimed that pricing significantly above
marginal cost was evidence of monopoly power, and that the discrepancy between actual
and theoretical monopoly price did not matter. But, any software is priced significantly
above marginal cost since marginal cost is zero, and the government has not yet sued
other software manufacturers on these grounds. Second, the government’s economic
witnesses claimed that the static monopoly model of the plaintiffs did not apply, but
offered no alternative model that could explain the difference between the actual price
and the monopoly price of the static model. Third, the government claimed that a large
29
It is very likely that the marginal price for the last unit sold to the same OEM was extremely low
since the 1995 consent decree allowed Microsoft to have quantity discounts but barred it from zero
marginal cost pricing.
30
The derivation of the monopoly price for Windows follows. Let pH be the price of the PC
hardware (everything except Windows) and let pW be the price of Windows. Assume that Windows is
installed on all PCs, i.e., that Microsoft has 100% market share. Since hardware and software are
combined in a ratio of 1:1, the demand of a PC with Windows is D(pH + pW). Profits of Microsoft from
Windows sales are:
ΠW = pW D(pH + pW) - FW
where FW is the fixed cost of developing Windows, and the marginal cost is negligible. Maximizing ΠW
implies marginal revenue equals marginal cost, i.e.,
D(pH + pW) + pW dD/dpW = 0 ⇔ 1 + [pW /(pH + pW)][(pH + pW)/D][dD/dpW] = 0 ⇔ pW /(pH + pW) = 1/|ε|,
or equivalently, the monopoly price of Windows is
pW = pH/(|ε| - 1),
where |ε| = -[(pH + pW)/D]/[dD/dpW] is the market elasticity of demand for PCs with Windows. If one
assumes that the average price of PC hardware is $1,800 and the elasticity is |ε| = 2, the monopoly price of
Windows is pW = $1,800. Even if one assumes a much higher elasticity, |ε| = 3, and a much lower average
price of PC hardware at $1,200, the monopoly price is $600, which is ten to twelve times the price charged
by Microsoft to OEMs.
17
market elasticity and revenues from complementary goods were sufficient to justify that
the actual price was equal to the monopoly price. But the required elasticities are very
large and inconsistent with the price to cost margins of hardware manufacturers.31
Why was the price of Windows low? For the early periods of each operating
system, one expects that the existence of network effects would prompt Microsoft to
charge a low price so that each platform becomes accepted by independent software
developers as well as users and the bandwagon gets rolling. This network-effects-based
theory does not explain why Microsoft did not increase the prices of each generation of
operating systems as each one matured. It also does not explain why Microsoft did not
increase significantly its price of Windows as it doubled its market share in the past four
years.
A variation of the network effects-induced low price theory gives a predatory
flavor to Microsoft’s strategy. In this view, Microsoft priced low to hook consumers and
generate network effects, while it planned to increase the price “in the future.” This
theory is particularly implausible because Microsoft dominates the PC desktop market
and has doubled its market share in recent years without increasing the price for
Windows while expanding its functionality. How long will Microsoft wait until it
increases the price of Windows? Some even claim that in a network industry, a firm can
practice predation without ever increasing the price, but just benefiting in the future from
the network effects. But such a strategy is indistinguishable from a truly competitive
strategy and cannot be considered predatory.
A number of other theories have been proposed to explain Microsoft’s pricing.
Some claim that the availability of an existing installed base of Windows constrained
Microsoft’s pricing. That could have been true if in fact Windows could easily and
legally be uninstalled by users. But Microsoft’s licensing requirements and the sheer
complexity of the uninstallation operation make it almost impossible for a user to
uninstall a Windows operating system that was installed by an OEM and move it to a
different (presumably new) computer. So the Windows installed base does not constrain
Windows pricing.
Others claim that the fact that software is durable constrains Windows pricing. It
is true that once Windows runs on an overwhelming market share of PCs, the substitute
for a new Windows computer is an old Windows computer. The fact that computers and
software are both durable makes this true. But very rapid technological change has
prompted consumers to buy new computers much faster than traditional obsolescence
rates would imply. Even doubling the price of Windows to OEMs would not have
implied a significant change to the final price of the combined computer and operating
system. So, durability of software and durability of computers, although a factor, is
31
In the Windows monopoly price formula pW = pH/(|ε| - 1), if the PC hardware price of $1,800 one
requires an elasticity of |ε| = 31 to get a Windows monopoly price of $60. A PC hardware price of $1,200
one requires an elasticity of |ε| = 21 to get a Windows monopoly price of $60. It is extremely unlikely that
that the market for PCs exhibits such large market elasticities.
18
unlikely to explain the huge difference between the actual price and the static monopoly
price.
Still others claim that the low price is implied by the very low cost of pirating
software. If that were the case, it would have prompted Microsoft to cut its much higher
prices of MS-Office and other software, since pirating takes the same effort and has the
same costs for pirates irrespective of the type of software being pirated. Moreover, the
control that Microsoft can exert on piracy of the operating system is much greater than on
piracy of applications. Therefore, although piracy could have been more of a problem if
the OS software was much more expensive, it is unlikely that the price of Windows is
low because of piracy considerations.
Finally some claim that Microsoft kept the price of Windows low because that
allowed Microsoft to charge more for complementary goods that it produces, such as the
Microsoft Office suite. There are three reasons that make this argument unlikely to be
correct. First, Microsoft also produces its most popular products, including the Microsoft
Office suite, for the Mac. If Microsoft kept the price of Windows low so that is sold MSOffice
for Windows at a high price, then the price of MS-Office for Macs should have
been lower than the price of MS-Office for Windows, which is factually incorrect.
Second, Windows has the ability to collect surplus from the whole assortment of
applications that run on top of it. Keeping Windows’ price artificially low would
subsidize not only MS-Office, but also the whole array of tens of thousands of Windows
applications that are not produced by Microsoft. Therefore, even if Microsoft had a
monopoly in the Office market, keeping the price of Windows low is definitely not the
optimal way to collect surplus. Third, even receiving large per unit revenues from
complementary goods cannot by itself explain the vast difference between the actual and
the monopoly price of Windows.32
Microsoft claimed that its low pricing was due to actual and potential
competition. Microsoft’s internal e-mails point to a real fear that the company would be
overtaken by the next innovator. Even if objectively it is difficult to see the big threat
from potential competitors, it is clear that Microsoft’s executives constantly felt the fear
of potential competition. On the pricing of Windows, I am inclined to believe
Microsoft’s view: Microsoft priced low because of the threat of competition. This means
that Microsoft believed that it could not price higher if it were to maintain its market
32
Suppose Microsoft receives net revenue RC sold from goods that are complementary to Windows
for every unit of Windows sold. Then its profits are
ΠW = (pW + RC)D(pH + pW) - FW .
Then the monopoly price of Windows is
pW = pH/(|ε| - 1) - RC .
Even if Microsoft makes a net revenue RC = $200 from complementary goods from every unit of
Windows sold, when pH = $1,800 and |ε| = 2, the monopoly price of Windows is pW = $1,600, and when
pH = $1,200 and |ε| = 3, the monopoly price of Windows is pW = $400, vastly above the actual price.
19
position. In essence, Microsoft pricing reveals that to a large extent its executives
believed that market conditions (most importantly potential competition) constrained
higher prices.
d. Exercise Of Market Power
The government claimed that Microsoft’s actions to exclude and marginalize
Netscape were sufficient to show that Microsoft possessed and exercised its monopoly
power. Essentially the government claimed that even if Microsoft did not exercise
monopoly power in pricing, it was sufficient to show that it exercised it in exclusionary
actions. But, most economists would agree that it is much more profitable to exercise
market power by increasing price than by raising the costs of rivals.33
If Microsoft
consistently sells at low prices, it loses a large amount of potential profits. It is very hard
to make a convincing argument that it is worth sacrificing these profits just to exclude a
future potential competitor. After all, even if the potential competitor is successful in
entering, it could only reduce Microsoft’s future profits, which are worth less today.
The government’s and the judge’s theory was that Microsoft exercised its
monopoly power by attempting to marginalize Netscape. The government and the judge
agree that Microsoft attempted to marginalize Netscape’s browser because Microsoft
feared that Netscape would become a rival platform to Windows. According to this point
of view, once Netscape became such a platform, applications would be written to run on
Netscape Navigator. Moreover, the bigger Netscape’s market share, the more likely it is
that ISVs would write applications for the Netscape platform; it follows that, in DOJ’s
view, competition in OSs is maximized when Netscape has a monopoly in browsers.
Since Netscape Navigator could run on many operating systems (not just Windows),
therefore, from this perspective, Netscape created a threat to Windows. In this view, all
actions to aggressively compete with Netscape were just attempts by Microsoft to defend
its monopoly in PC operating systems.
This is an interesting theoretical argument, but it has both theoretical and
empirical deficiencies. First, in winner-take-most competition, DOJ’s analysis fails to
distinguish between profit sacrifices to keep out potential rivals and competition where a
company keeps its prices low in order to win consumer business from actual and potential
rivals. Second, although at some point in time Marc Andreesen, Netscape’s CTO,
claimed that Netscape would create the capabilities of running applications, there was
never a consistent effort by Netscape to create a comprehensive set of Applications
Product Interfaces (“APIs”) that would support typical applications that now run under
Windows. Moreover, Netscape hardly put sufficient effort and resources in reaching
independent applications developers to write for such a platform. So, although
potentially Netscape could have created such a platform, it was very far away from that
goal, and the probability of succeeding against the array of Windows applications was
low.
33
The exception is regulated industries where a firm is not free to increase price.
20
Third, Netscape, even if successful, is likely to have been limited to Internetbased
applications. The performance of such applications is dependent on the Internet
working efficiently and at top speed. Actual performance would also potentially be
hampered by excessive loads on the CPUs of servers where the applications would
actually run. Essentially the model of Netscape as competitor of operating systems
required a number of contingencies to be met, and therefore faced a very significant
uncertainty. If instead applications were to run on the PC itself (on top of Netscape) they
would need to interface to some extent with the underlying operating system. If this
system was Windows, then it would be unlikely that it would accommodate perfectly its
direct competitor.
Fourth, Microsoft’s actions clearly could have had competitive justifications.
Microsoft’s management saw that the Internet was taking off and wanted to have a
dominant position in it.
DOJ and the District Court Judge also agreed that Microsoft sabotaged Sun’s Java
operating language so that Java would not become a “universal language” in the sense
that applications written for it could be run in any operating system. Over the years, the
computer industry has gone through a number of such pipe dream “universal languages.”
They have all failed for a very simple reason: a “universal language” cannot be too
specialized -- after all it has to be recognizable by diverse operating systems. But
efficiency and speed in program execution requires specialization. Universal languages
eventually fail because they get more and more efficient for particular operating systems
and thereby their universality fails.34
Microsoft created an implementation of Java for Windows that was more efficient
than the original Java by Sun. Microsoft’s implementation (that is, Microsoft’s Java
virtual machine) allowed for all programs written for the original (“pure”) Java to be run
on it. Thus, it preserved backward compatibility with the original Java that ran on all
operating systems. Because of that, Microsoft’s actions were not anti-competitive.
Microsoft claims that Sun’s non-OS-specific Java language was inefficient and slow, and
that Microsoft improved it. This is a reasonable pro-competitive justification for
Microsoft’s actions. The fact that Microsoft’s Java implementation also runs programs
that do not run in other Java implementations is not cause for anti-competitive concern,
especially since Microsoft has published the function calls its Java makes to Windows.
Competing operating systems could implement these calls.35
A key economic witness of DOJ, Prof. Frank Fisher of MIT, proposed in pre-trial
filings and supported during the trial a novel theory of predation. The traditional theory
of predation requires that a product is sold below incremental (or “avoidable”) cost for a
34
The fate of human languages is similar. As each language becomes more efficient for a certain
group it gets more specialized and cannot easily be understood by the rest of the world.
35
This discussion is limited to antitrust considerations and does not discuss the contractual dispute
between Microsoft and Sun.
21
period of time until competitors are driven out of business; then the monopolist increases
prices and reaps its monopoly profit.36
Instead Prof. Fisher calls an action predatory if it
would not have been profitable (and therefore not undertaken) in the long run unless the
predator was taking into consideration profits arising out of the negative impact of the
practice on competition.37
According to this definition of predation, predation can occur
even when the predator sets price above incremental cost, or even above average cost.
All that is required for predation, according to Prof. Fisher, is for a company to take some
(presumably very aggressive) action that would not be rewarding unless competitors are
hurt. If this definition is to be adopted, a large array of actions that are normal business
practice and which benefit consumers are going to be characterized as predatory and
therefore illegal.
In antitrust economics, it is widely believed that rules should be fashioned to
maximized social surplus. The Fisher rule fails to do so. It creates an artificial cushion
of profits. If an aggressive competitor dares to take away some of these profits, he is
named a predator. Moreover, since the profit cushion is created without regard to cost, it
can be at any status quo level. Instead of promoting competition and efficiency, the
Fisher predation theory could protect slack and monopoly rents. The judge erred to the
extent that he based his ruling on this incorrect theory.
To understand better Prof. Fisher’s definition of predation, consider the following
example. Suppose that in a two-firm industry, firm 1 has constant average unit cost equal
of $10, while firm 2 has constant average unit cost of $15. An “accommodation strategy”
of firm 1 would be to charge a price above $15 (say $16), so that the market is shared
between firms 1 and 2. Suppose that firm 1 plays an “aggressive strategy” and charges a
unit price of $14. Such an action results in a profit for firm 1 of $4 for every unit sold in
the market. These profits are likely to be higher than the profits firm 1 would have
received in the accommodation strategy of $16. The accommodation strategy gives a
profit of $6 per unit sold by firm 1 and, since firm 1 sells only half the output, its profit is
$3 for every unit sold in the market. If the market is not extremely elastic, firm 1 makes
more profits if it uses the aggressive strategy, which drives firm 2 out of business.
Applying Prof. Fisher’s predation definition, firm 1 is a predator because its aggressive
action increased its monopoly rent. Moreover, in comparison with the “accommodation
strategy,” the aggressive strategy would have not been taken if it were not trying to
increase its monopoly rent.
If the “aggressive strategy” is considered predatory and therefore illegal according
to Prof. Fisher’s criterion and firm 1 uses the “accommodation strategy” instead, total and
consumers’ surplus will decrease. Consumers will face a higher price of $16 rather than
$14 and will therefore be worse off. Moreover, firm 2 will be producing half the units
sold at a cost that is higher by $5 than the most efficient production cost, thereby
reducing total surplus. It is evident that application of Prof. Fisher’s predation criterion
36
See, for example, Areeda and Kaplow (1997), p. 512, 514.
37
Specifically Prof. Fisher’s standard is that an action is anti-competitive when such an action
“would not be profit maximizing except for the monopoly rents earned or protected by the action.”
22
can result in lower consumers and total surplus. Moreover, the application of this
criterion shields the less efficient firm from competition and guarantees its rents.
8. On Tying Of Windows With Internet Explorer
As we have seen, the District Court ruled that Microsoft had monopoly power in
the operating systems market. Judge Jackson further ruled that the Internet browser
market was separate from the market for operating systems, and that technological and
contractual tying of Internet Explorer with Windows raised prices and hurt consumers.
The judge ruled that consumers do not want IE even free of charge because it burdens the
OS with memory and overhead requirements, and it consumes a significant amount of
hard disk space.
There is a major contradiction in the District Court’s decision. In the Sherman ¶2
part of the case, browsers are in the same market as operating systems, but in the
Sherman ¶1 part of the case, browsers and operating systems are in different markets. On
the one hand, in the tying part of the case, the Judge ruled that Internet browsers are in a
separate market from PC operating systems. On the other hand, in the monopolization
part of the case, the Judge ruled that Microsoft attempted to marginalize the Netscape
browser to defend its Windows monopoly because the Netscape browser was (or was
becoming) a close substitute for Windows. If Internet browsers are in a separate market
from the operating systems market, then Microsoft’s actions in the browser market in the
monopolization part of the case cannot be interpreted as defending Microsoft’s
monopoly. If instead the browser is in the same market as Windows, there cannot be
illegal tying.
Microsoft claimed that IE was giving crucial functionality to its operating system
and had every right to include it in Windows under the terms of the 1995 consent decree
which allowed addition of functions and features to Windows. Although this is a legal
matter, I believe that the consent decree’s language is plain enough to allow one to
conclude that Microsoft’s view is correct.
But why did Microsoft produce a browser and why did it then give it away?
Essentially the question is: what is the appropriate behavior in a complementary goods
market for a firm that has a dominant position in one market? That is, does it make sense
from a competitive point of view that Microsoft was giving away the browser at no
charge, or was it an anti-competitive act? To understand this, we first outline some
thoughts about extracting value from complementary goods chains, where one good has
close complements that are used in conjunction with it.38
First, a firm in one market is better off and can extract more value when the
complementary markets in which it does not participate are more competitive.
Conversely, the more monopolized a market is, the less value remains for the
38
This discussion follows closely Economides (1999).
23
complementary markets. Second, keep in mind that value can be extracted only once
from the chain of components. If a firm can monopolize one component, and all other
complementary component markets are perfectly competitive, the firm gains nothing by
attempting to monopolize the complementary component markets. The value of a good
or service cannot be extracted more than once, no matter how many components the good
is broken into. But, a firm that monopolizes a component has an incentive to enter and
compete hard in a market for a complementary component if the complementary
component market is not perfectly competitive, because by entering such a market, the
firm can capture rents that it was losing from its original market. Third, more generally,
a firm that participates in a market that is not perfectly competitive has a stronger
incentive to enter a (not perfectly competitive) market for a complementary component
than a firm that does not participate in any other market. Otherwise put, when
complementary component markets are not perfectly competitive, there are strong
incentives for the same firms to enter more than one of these markets. Fourth, when a
component, say component A, is used together with many other complementary
components, say B1, ..., Bn, to produce n composite goods of varying market values,
competitors in the market for A will tend to enter the markets for the complementary
components that, in combination with A, have the highest market value.
The entry of Microsoft in the browser market can be seen in the context of
competition and multi-market participation in markets for complementary components.
As long as the market for browsers was competitive and was shared by a number of
browsers, and as long as the Internet was a small academic market, Microsoft had no
significant business interest in it. When the browser market became dominated by
Netscape, and, simultaneously, the Internet market appeared to be significantly larger,
Microsoft integrated browser features in Windows and significantly improved its
browser. Again, there were two reasons for this. First, Netscape had a dominant position
in the browser market, thereby taking away from Microsoft’s operating system profits to
the extent that Windows was used together with the Navigator. Second, as the markets
for Internet applications and electronic commerce exploded, the potential loss to
Microsoft from not having a top browser increased significantly. These reasons were
sufficient to drive Microsoft to produce a browser.
Was the free distribution of IE anti-competitive? Clearly, Microsoft had a procompetitive
incentive to freely distribute IE since that would stimulate demand for the
Windows platform. Were there anti-competitive effects? If one assumes (as DOJ does in
the tying part of the case) that the two markets (operating systems and browsers) are
separate and therefore tying occurred, the crucial question is whether tying was anticompetitive.
In particular, was the incremental increase in the price of Windows when IE
was bundled with it larger than the price increase justified by the value of functionality
that IE added? This is the test that the Court should have conducted, and it failed to do
so. My view is that it is very hard to prove that a quality-adjusted new version of
Windows without IE would not have had a higher hedonic price than old Windows. That
is, a modest increase in the market price of Windows was likely to be justified by the
enhancement of features of Windows even without the inclusion of IE. Thus, if the Court
had performed this test, I believe it is likely that it would have found that adding IE
24
functionality to new versions of Windows and distributing IE free of charge for older
versions of Windows and for other operating systems did not harm consumers.
9. On Attempting To Monopolize The Browser Market
The District Court Judge found that Microsoft lost money while trying to gain
market share from Netscape, in its attempt to monopolize the browser market by
leveraging its monopoly of the OS market. The Judge also found that Microsoft failed in
this attempt.
It is well understood that the potential surplus of a consumer for a unit of a good
can only be extracted once. If Microsoft is able to extract the surplus of a consumer that
buys an Internet browser through Windows, it has no incentive to try to monopolize the
market for browsers, as long as the market for browsers is competitive. To the extent that
the market for browsers is monopolized, it can take away from the surplus appropriated
by Windows, and then Microsoft may have an incentive to expand its market share in
browsers. There is significant evidence that the market for browsers was originally
monopolized by Netscape. Then the inroad of Microsoft into browsers replaced a
monopoly with duopoly. And, it brought the price to zero and intensified research and
development in browsers to the great benefit of consumers. It is quite ironic that,
although the inroad of Microsoft into browsers was greatly beneficial to consumers and
broke the Netscape monopoly, the government and the judge found Microsoft’s actions
anti-competitive.
10. Other Potentially Anti-competitive Arrangements
The court proceeding spent a significant amount of time on potentially anticompetitive
actions of Microsoft in its dealings with buyers, content providers and others.
An examination of the various arrangements and the facts of each one of them is
impossible in this article due to of lack of space. Broadly speaking, the potentially anticompetitive
actions were of three kinds.
1. Contracts with OEMs on the distribution of Netscape and IE, and the lack of
control by OEMs of the opening screen (“first screen”) of a new PC.
2. Contracts with ISPs (such as AOL) on the distribution of Netscape and IE
3. Control by Microsoft or OEMs of “active channels” to be provided to Internet
Content Providers (such as Disney)
The potentially anti-competitive effects of such contracts have to be judged on a case by
case basis.
11. Effects On Consumers
In principle, there are three ways that consumers could be hurt by potentially anticompetitive
actions. First, consumers may be hurt because these actions increase prices.
Second, consumers may be hurt because these actions may limit their choices in terms of
25
variety and quality. Third, these actions may limit innovative activity, thereby hurting
future consumers. Any harm in any of these three dimensions should be evaluated and
balanced with any benefit of these actions in other dimensions. It is customary in
antitrust cases to value the benefits to consumers and balance them with the burden of
anti-competitive actions. The District Court failed to do this task. The Court did not
attempt to quantify the benefits to consumers from giving away Internet Explorer and the
intensification of competition that resulted. Also, the Court did not try to assign a
monetary value to the losses to consumers resulting from the anti-competitive actions for
which it found Microsoft liable.
Consumers have directly benefited from the free distribution of Internet Explorer
as well as its bundling and tight integration with Windows. When Microsoft started to
seriously compete with Netscape in the Internet browser market, Netscape -- essentially
the sole provider of Internet browser software -- charged non-academic users $40-50 to
use its browser.39
Microsoft, by contrast, gave its Internet browser away. Netscape
responded to the introduction of IE version 4 that many held of superior or comparable
quality to Netscape by giving its browser away as well. Today, with at least 100 million
browsers installed in the United States, Microsoft’s actions have created a benefit of at
least $4 to $5 billion to American consumers. And, since Microsoft’s actions intensified
competition, which in turn produced higher quality browsers, they provided further
benefits to consumers.
The District Court Judge ruled that the prices of Windows 95 and 98 were too
high, based on internal Microsoft e-mails that discussed a range of prices considered by
the company for pricing Windows. However, consumers may have directly benefited
from the relatively low price of Windows. Since the marginal cost of Windows is almost
zero, clearly Microsoft had a wide range of prices it could have used. The static
monopoly price for Windows was thirty to forty times higher. Microsoft’s operating
system, for which computer manufacturers pay $40-60 per copy, is cheap compared to
the historical and current prices of other operating systems. For example, in the late
1980s, IBM sold OS/2 (which ran much fewer applications than Windows) for hundreds
of dollars. Some Linux packages -- essentially add-ons to the free Linux source code –
currently sell for $150, and run far fewer applications than Windows does. These price
discrepancies highlight a huge contradiction in the government’s case and in the judge’s
findings of fact. If Microsoft were a true malevolent monopoly unconstrained by
potential competition, it would charge far more for Windows than it does. The annual
consumer benefits from Windows’ relatively low price may be many billions of dollars.
The District Court judge ruled that Microsoft’s action of distributing Internet
Explorer at no charge “increased [the] general familiarity with the Internet and reduced
the cost to the public of gaining access to it,” “gave Netscape an incentive to improve
Navigator’s quality,” and benefited consumers since it “compelled Netscape to stop
39
Netscape was distributed free of charge to academic and governmental institutions. Netscape was
also distributed free of charge to all in its introductory stage.
26
charging for Navigator.”40
On the other hand, the District Court judge ruled that
consumers were hurt in various ways by Microsoft’s anti-competitive actions. The judge
ruled that consumers were hurt41
because they could not get Windows without a browser
since Microsoft “forced OEMs to ignore consumer demand for a browserless version of
Windows,” and that the inclusion of Internet Explorer led to “degraded system
performance, and restricted memory.”
I believe that the Judge erred in ruling that free distribution of IE significantly
harmed consumers by degrading performance and restricting available computer
memory. First, because IE has been integrated into Windows, it is very hard to judge
how a computer would work without it, how much memory it would use, how quickly it
would run, etc. One would have to construct a computer with some other new
component of the OS taking over the functions that IE performs now in Windows, and
then test it. The Court failed to do so, and cannot conclusively rule on this issue.
Second, the Internet-related tasks of operating systems are in very high demand as the
Internet is presently in an exponential expansion path. The number of consumers who
have been harmed in the way suggested by the Judge is likely to be very small, and in an
age of cheap computers, memory, and hard drive capacity, their losses could not be very
large.
Netscape and Internet Explorer ended up being very similar products in terms of
their functionality. Most computer magazines rate recent versions of IE above
Netscape’s. There does not seem to be any loss in variety or quality by the domination of
the browser market by IE. Windows-based computers can run both IE and Navigator
simultaneously, and users are not forced to choose one to the exclusion of the other. One
could use one browser for some tasks and another for others.
The court ignored the issue of compatibility which was probably the most central
issue in the case. Backward and forward compatibility are crucial for software markets,
and, as explained earlier, are the source of network effects. Microsoft provides de facto
compatibility through its Windows operating systems. Compatibility is an important
benefit to society because it is the source of network effects. Compatibility and its
benefits could be quickly eliminated or significantly reduced if Microsoft is broken into
competing pieces. Microsoft did not forcefully present to the Court the benefits to
society of its dominant position arising from the de facto compatibility it created and the
resulting network effects.
Finally on the issue of innovation, economists’ opinions are split on whether
monopoly or competition would create more innovation. Economists’ opinions are also
split on whether vertically integrated or independent companies create more innovation.
The government’s economic witnesses did not provide any specific theory to support the
40
Findings of Fact, United States v. Microsoft, Civil Action No. 98-1232 (TPJ) & State of New
York v. Microsoft Civil Action No. 98-1233 (TPJ), thereafter (“FOF”), at paragraph 408.
41
FOF at paragraph 410.
27
claim that Microsoft was damaging the innovation process. The theory that general fear
of Microsoft kept competitors out of the market or without available capital is
contradicted by the abundant support by financial markets of the successful entry of a
number of Linux variants.
12. Remedies
a. Remedies As Imposed By The District Court
The issue of remedies was from the very beginning the most difficult part of the
litigation. Many observers that believed that Microsoft was liable had expressed
privately great consternation in finding an appropriate remedy. Given the variety of
remarks of Judge Jackson during the trial, there was no doubt that the District Court
would have found Microsoft liable. Even before the trial started, it seemed that the
remedies determination would be the most crucial and most interesting element of the
process. It was widely expected that the remedies phase would be a new full scale trial
with significant input by economists.
Instead, Judge Jackson decided on remedies in summary fashion without any
examination of the proposals of all sides. He was quoted shortly afterwards arguing that
since the plaintiffs won the liability phase, the plaintiffs should have the right to
determine the remedies.
In the last week of March 2000, the New York Times published the terms of the
Microsoft and DOJ negotiations conducted under Judge Posner’s supervision. The
published information implied that DOJ and Microsoft were not far from an agreement,
but that some States that were part of the litigation disagreed with the settlement terms.
The terms of the final draft of the proposal of DOJ were:42
1. Microsoft would create a pricing schedule that would apply to all buyers, so that
prices would not be conditioned on other Microsoft products that a buyer buys.
The schedule would allow for different prices for different quantities.
2. Microsoft would not be allowed to have exclusive contracts that do not allow the
other party to use, display, or feature its opponents products.
3. Microsoft would be required to share technical information without
discrimination among the potential recipients of this information.
4. Microsoft would be required to disclose the application interfaces (APIs) that link
applications to Windows.
5. Microsoft cannot increase the price of old versions of Windows when new
versions are released. Microsoft would be forced to sell the old version at a
constant price for three years after a new release.
6. Microsoft would produce a Windows version without IE. Computer
manufacturers would be allowed to license some part of the Windows code so that
they could change the opening screen, and choose the default browser
42
New York Times, April 2, 2000. See also Brinkley and Lohr (2000), page 286-7.
28
7. Tying by contract would be prohibited, but Microsoft would be allowed to
integrate functions, applications, and features in its products.43
This list of conditions seemed a reasonable basis on which a deal with Microsoft
could be struck. Since these were the government’s proposals, one expects that the final
negotiated settlement would have been somewhat less harsh on Microsoft. It was widely
reported that some States officials considered these remedies as less harsh than they
wished for. These States officials communicated to Judge Posner that they would not
sign such a deal. Judge Posner was forced to declare the negotiations a failure.
Given the proposals that the government offered in the negotiations, there was a
wide expectation that DOJ would demand more or less the same terms in the remedies
phase.44
Instead, DOJ asked for a much more radical step: a breakup of Microsoft.
The plaintiffs remedies proposal as adopted by the Judge imposed a breakup of
Microsoft into two “Baby Bills,”45
an operating systems company which would inherit all
the operating systems software, and an “applications” company with all the remaining
software assets. Cash and securities holdings of other companies held by Microsoft
would be split between the resulting entities. Bill Gates and other officers / shareholders
of the company would not be allowed to hold executive and ownership positions in both
of the resulting companies
The District Court ruling also imposed interim conduct restrictions on Microsoft.
These restrictions, to last three years, from the time of the breakup were:
1. Microsoft would create a pricing schedule that would apply to all buyers, so that
price would not be conditioned on the sale of other Microsoft products.
2. Microsoft would not be allowed to have exclusive contracts that do not allow the
other party to use, display, or feature its opponents products.
3. APIs and other technical information of Windows should be shared with outsiders
as it is shared within Microsoft.
4. Microsoft is not allowed to take actions against manufacturers who feature
competitors’ software.
5. Microsoft will allow OEMs to alter Windows in significant ways.
6. Microsoft is not allowed to design Windows to disable or compromise rivals’
products.
43
This seems identical to the provision of the 1995 decree which the DOJ claimed at the beginning
of this lawsuit was not restricting Microsoft enough.
44
In fact, Prof. Harry First who was acting as Assistant Attorney General for Antitrust for the State
of New York was quoted in the New York Times (April 30, 2000) saying that he and other officials of the
States were expecting DOJ to propose remedies similar to DOJ’s proposals at the negotiations, and were
very surprised when Assistant Attorney General Joel Klein informed them a few days before the filing that
DOJ would propose a breakup. Also see Brinkley and Lohr (2000), pages 306-7.
45
This is a word play on “Baby Bells” that came out of AT&T and the first name of the CEO of
Microsoft, Bill Gates.
29
These conditions are similar but more restrictive than the ones proposed by the
government in the settlement talks at the end of March 2000. At that point in time the
government was satisfied that these restrictions, together with a ban on tying and a
requirement that Microsoft would produce a version of Windows without IE, were
sufficient to settle the case. DOJ did not produce any significant reason that made it
change its position. Moreover, DOJ did not produce any convincing argument that the
proposed breakup would be more efficient or effective than the conduct restrictions that it
had proposed earlier.
In arguing for the break-up, the government put forward a number of reasons.
But, since there was only a summary formal hearing on remedies, there was no chance
for both the government’s and Microsoft’s cases on remedies to be discussed and
evaluated. The government and the judge have stated (formally and informally) the
following arguments for a breakup:
1. That it considered the repeated violations of antitrust law by Microsoft as an
indication that Microsoft would not follow any conduct or contractual restrictions;
in fact, in some informal remarks, government officials believe that they were
“tricked” by Microsoft in settling the 1995 case with terms that Microsoft was
able to exploit;
2. That the lack of remorse by Microsoft’s executives was a clear indication that
Microsoft “could not be trusted” to implement any other remedy;
3. That the breakup was a “surgical cut” that would create the least interference with
business;
4. AT&T and the rest of the telecommunications industry benefited from AT&T’s
breakup, and so should Microsoft and the software industry; after all both
industries have network effects;
5. The breakup eliminates the incentive for vertical foreclosure; and
6. The breakup reduces the “applications barrier to entry” since now the applications
company might write popular Microsoft applications (such as MS-Office) for
other platforms.
I believe that the government has failed to show that the proposed breakup is the
appropriate remedy. DOJ had to show that conduct remedies would not work and it
failed to do so. DOJ has not performed the appropriate cost-benefit analysis to show that
conduct remedies are not sufficient and that a breakup is necessary. None of the
economists affidavits in the remedies phase even come close to discussing and evaluating
alternatives. As discussed earlier, the government also failed to justify why it was ready
to compromise a few weeks earlier on behavioral remedies but now it claimed that
structural remedies were necessary.
The first argument of the government does not stand to reason. The 1995 case
was settled with a decree that explicitly stated that Microsoft can include in its operating
system any additional functionality. It is reasonable that Microsoft (or any observer,
including DOJ) would believe, given the 1995 consent decree, that adding browser
functionality to Windows does not violate the consent decree. This, of course, does not
mean that adding such functionality does not violate antitrust law in general, but it puts to
30
its death the idea that the government was tricked by Microsoft. Of course, it stands to
reason that Microsoft understood much more clearly what the degree said than DOJ since
Microsoft understood the business better and could predict what functions could be
included in the operating system. However, the fact that companies and antitrust
enforcers often have an asymmetry of information is very common and expected, and
cannot be considered a “trick,” or a reason not to enter into agreements between antitrust
authorities and companies.
The second argument of the plaintiffs makes no sense at all. Antitrust
enforcement is not an emotional tug of war in which the egos of the either the plaintiffs
or the defendants need to be satisfied. The show of remorse or lack thereof by Microsoft
executives could not possibly define the remedy. I find it hard to believe that the judge
would really find a different remedy appropriate if enough Microsoft executives simply
showed public remorse. Moreover, Microsoft, like any other defendant, has a right to
appeal. Belief that it will prevail on appeal in a civil case is hardly worthy of
punishment.
The third argument, that the breakup is a surgical cut and therefore will disrupt
the industry the least is countered by the facts. A breakup of Microsoft, if it is finally
ordered at the end of the appeals process, would, practically speaking, eliminate
Microsoft as a flexible and formidable competitor. The wholehearted endorsement of the
breakup by Microsoft’s competitors in the servers and backoffice competitors (who were
not alleged to have been damaged by the Windows monopoly but will greatly benefit
from the confusion and disruption created by a Microsoft breakup) is evidence that the
breakup is one of the most disruptive possible outcomes.
The fourth argument, that since AT&T’s 1982 breakup was successful so would
Microsoft’s, is incorrect. AT&T was divided into the long-distance company (AT&T),
and seven regional operating companies, each of which remained a regulated local
telecommunications monopoly until 1996. The destruction of AT&T’s long-distance
monopoly encouraged competition, which brought sharply lower prices and immense
consumer benefits.46
There are a number of key differences between the two companies
and their competitive situations. And these differences make it very likely that a
Microsoft breakup, besides harming Microsoft, would harm consumers and the computer
industry.
In 1981, AT&T was a 100-year-old regulated monopoly with many layers of
management. For historical reasons, the local phone companies within the old AT&T,
such as New York Telephone, were managed separately from the “long lines” division.
Thus, it was not difficult to separate the divisions since they functioned on many levels as
separate companies. AT&T also had an abundance of managers to help cope with the
46
At the same time the AT&T breakup did not introduce competition at the local exchange level,
and the Regional Bell Operating Companies (“RBOCs”) were allowed to monopolize local
telecommunications services as well as access to long distance services. The success of competition in
long distance has been hampered by the continuing monopoly of the local exchange, four years after the
Telecommunications Act of 1996 was supposed to open the local exchange to competition.
31
breakup. By contrast, Microsoft is a young, entrepreneurial company run by very few top
executives (about 25), and its divisions are very fluid. While this has made Microsoft
one of the most efficient and successful companies around, it also means that a break-up
would pose significant managerial problems and severely reduce the company’s
flexibility. Finally, AT&T was a regulated utility and regulation guaranteed that the
companies emerging from the breakup stayed interconnected. In contrast, the Microsoft
breakup is likely to lead to incompatibilities and further loss of efficiency.
DOJ’s two-way breakup plan was premised on the hope that an autonomous
applications company would create a new operating system to compete with Windows.
But more than 70,000 applications run on Windows, creating what the government calls
“the applications barrier to entry” in the operating-system market. However capable the
new applications company, it still wouldn’t be able to single-handedly create a successful
rival operating system. Separately, even with a new applications company’s support,
Microsoft’s biggest operating-system competitor, Linux, is unlikely to become a serious
desktop threat to Windows.
In my opinion, the breakup as proposed by the government and imposed by the
District Court will have detrimental effects. First, the breakup is likely to result in higher
prices. If DOJ is correct and Microsoft kept its OS prices low so that it can exercise its
monopoly power in the adjacent browser market, the post-breakup Baby Bill that inherits
the operating systems will have no incentive to keep the price low. The OS Baby Bill
will no longer have the incentive to disadvantage any applications companies. Thus, if
DOJ’s theories are correct, the OS Baby Bill will now exercise the monopoly power it
has and raise the price of the operating system to the detriment of consumers. If DOJ is
correct and Microsoft has significant monopoly power because of the “applications
barrier to entry,” higher prices will be the direct result of the breakup. Second, as
explained earlier, the breakup is likely to eliminate the efficiencies that make Microsoft a
flexible and formidable competitor.
The breakup is likely to temporarily eliminate the incentive for interference from
OSs to applications and vice versa. Of course, the same could have been accomplished
by conduct restrictions without the cost and the disruption of the breakup. Moreover, the
District Court did not impose permanent restrictions on the post-breakup functions of the
companies. The OS and the applications Baby Bills may enter into each other’s business
soon after the breakup. It is very likely that a few years after the breakup, one of the
resulting companies will dominate both markets.
b. Other Remedies Proposals
A number of other remedies proposal have been discussed in the literature, but
unfortunately were not discussed in the District Court’s summary remedies proceedings.
If the liability conviction is upheld by the Court of Appeals or the Supreme Court, it is
likely that the issue of remedies will be re-opened and a full hearing on remedies may
take place.
32
In principle, remedies fall into two broad classes. First there are those that attack
the business side of Microsoft by restricting its contracts or splitting the company along
various lines of business. The second broad class attacks the control over the technology
by either forcing Microsoft to release or license the source code, or to disclose
proprietary information (APIs). The remedies can also be divided into those that affect
conduct and those that affect structure. For example, a restriction on contracts is a
conduct remedy, while a breakup is a structural remedy.
Provided that the liability conviction is upheld, four alternatives have been
proposed. One remedy proposal is even more extreme than the one imposed by the
judge. This extreme proposal would break up Microsoft into three identical Baby Bills,
with each company acquiring the source code of all the programs that Microsoft currently
sells, and one third of its employees.47
This “horizontal remedy” is sometimes presented
in combination with the “vertical remedy” imposed by the judge. In this “hybrid
remedy,” first Microsoft is broken into two or three companies according to the type of
program produced, and then the operating systems company is broken into three parts
creating four of five companies altogether.
Besides the loss of flexibility that any breakup would create, a horizontal or
hybrid breakup would also produce significant incompatibilities with harmful effects to
computer users, applications writers, and Microsoft shareholders. Baby Bills coming out
of a horizontal or a hybrid breakup will have incentives to create incompatible versions of
Windows for two reasons. First, Baby Bills will try to differentiate their operating
systems to avoid strong competition, leading to small price-cost margins. This is true
even in industries without network externalities and has been well established in the
economics literature on product differentiation.48
Second, Baby Bills will try to make
their operating systems incompatible with each other in a race to become the dominant
OS, since the dominant firm receives the lion’s share of profits in a winner-takes-most
world. This is established in the network economics literature.49
Differentiating the
operating systems by Baby Bills would inevitably reduce the range of software that
would be compatible with each user’s computer. As a consequence, consumers’ surplus
would decrease. The emerging incompatibilities would be a huge headache for both
independent applications writers and corporate IT departments. Such incompatibilities
would also hurt shareholders, since the combined value of the resulting Baby Bills will be
smaller than that of the original Microsoft.
Another remedy proposal is auctioning the Windows source code. Given the
fluctuating stock market value of Microsoft, Windows source code may be worth as
much as $200 billion. No company can bid that much cash in an auction. (Practically
speaking, only a handful of foreign governments could). This implies that the source
code of Windows would be sold forcibly at a small fraction of its worth – and that would
47
See Litan et al. (2000).
48
See, among others, D’Aspremont et al. (1979), Shaked and Sutton (1982), and Economides
(1984).
49
See Economides and Flyer (1998), and Economides (2001).
33
severely reduce the value of shareholders’ equity. Auctioning the Windows code would
not only effectively confiscate Microsoft’s intellectual property, it would also seriously
reduce the incentive for innovation not only for Microsoft but for all potential innovators.
Moreover, source code evolves. Over time, different firms will add and alter the
Windows code. Soon, incompatibilities will arise, with all the negative consequences of
diminution of network effects described earlier.
A second remedy proposal would be to impose on Microsoft to disclose some of
the so-called “APIs” (lines of software code that define interfaces between applications)
that permit it to include Internet Explorer in the operating system. Microsoft routinely
discloses APIs that hook applications to the operating system and allow for
interoperability. Currently, it does not disclose APIs that tie together parts of the
Windows operating system, which includes Internet Explorer. If Microsoft were to
disclose the APIs that hook Internet Explorer to other parts of the operating system,
Netscape (or any other browser) could have the same interoperability with Windows.50
A third remedy proposal – and in my mind, the best -- would be to consider
imposing various restrictions on the contracts that Microsoft can write with sellers of
complementary goods and with competitors. This is a likely remedy that is easy to tailor
according to the violation. This remedy – possibly combined with requiring Microsoft to
disclose certain APIs -- should be sufficient to guarantee that Microsoft will be precluded
from taking future anti-competitive actions. At the same time, it would preserve the
managerial and other benefits that have made Microsoft one of the most successful and
most profitable companies ever.
Some have proposed regulation of the operating systems part of Microsoft as a
way to guarantee that there would be no leverage of market power. There are very
significant reasons to avoid regulation at this stage. First, regulation is inappropriate for
an industry that evolves quickly. Regulation requires a stable product. For example,
regulation of AT&T did not occur until 35 years had passed after the invention of the
telephone, despite the fact that (i) telecommunications had much stronger network effects
than software; and (ii) during its first 40 years, AT&T took much more aggressive actions
than Microsoft, including refusal to deal and refusal to interconnect with competitors.
Second, regulation could stagnate innovation. Third, in an industry with fast technical
progress, regulation can be used by the regulated companies to keep prices high, as
exemplified by telecommunications regulation.
13. Concluding Remarks
The immediate future of the Microsoft case brings the appeals process (at the DC
Circuit Court of Appeals or the Supreme Court) well into the presidency of George W.
Bush. It is almost certain that the appeals court will reverse the tying decision and will
find Microsoft not liable on tying. In my opinion, it is likely that the major parts of the
50
This is a disclosure of a few APIs that hook IE to Windows. It does not imply a disclosure of a
large number of APIs, which would be almost equivalent to disclosure of the Windows source code.
34
rest of the District Court’s decision would also be reversed. And, it is most likely that the
federal part of the case will be settled after the Court of Appeals decision. Microsoft will
have very significant legal fees and related costs, as it also faces a large number of civil
antitrust suits by class action plaintiffs. The breakup would be disastrous, but, in my
opinion, is extremely unlikely to be upheld. The biggest loss to Microsoft is the
continuous antitrust scrutiny that does not allow it to make significant acquisitions in
telecommunications and the Internet in the United States during the period of intense
antitrust scrutiny.
The antitrust trial has put a tremendous pressure on Microsoft. If the company
was not controlled by Bill Gates and employees, it is very likely that, in the short-term
interest of the shareholders, Microsoft would have compromised with DOJ, even though
this is likely to have harmed the long-term interests of shareholders. As things stand in
August 2000, the stock price of Microsoft is down about 40 percent because of the
continuing antitrust uncertainty.
Regardless of the final outcome, the effects of U.S. v. Microsoft are likely to be
felt for a long time. If a breakup actually occurs, it is likely to impose the dark shadow of
radical antitrust intervention on the whole computer industry. There are many firms in
the computer sector that have a dominant position in their respective markets. I would
expect, if on appeal DOJ wins big on Microsoft and if DOJ takes a principled approach,
antitrust suits against Yahoo, AOL, and other pioneers of the New Economy will not be
far behind.
35
14. Bibliography
Amicus Curiae Brief Of Professor Lawrence Lessig, United States v. Microsoft, Civil
Action No. 98-1232 (TPJ).
Areeda, Phillip and Louis Kaplow (1997), Antitrust Analysis, Aspen Law & Business.
Brinkley, Joel and Steve Lohr (2000), U.S. v. Microsoft, McGraw Hill.
Cass, Ronald A., (1999), “Copyright, Licensing, And The ‘First Screen’,” mimeo.,
Boston University, School of Law.
Cass, Ronald A., and Keith N. Hylton, (1999), “Preserving Competition: Economics
Analysis, Legal Standards And Microsoft,” Boston University, School of Law,
Law & Economics Working paper no. 99-1.
Conclusions Of Law, United States v. Microsoft, Civil Action No. 98-1232 (TPJ) & State
of New York v. Microsoft Civil Action No. 98-1233 (TPJ).
Davis, Steven J., Jack MacCrisken, and Kevin M. Murphy, (1998), Integrating New
Features Into The PC Operating System: Benefits, Timing, And Effects On
Innovation,” mimeo.
D’Aspremont, Claude, Jaskold-Gabszewicz, Jean, and Thisse, Jacques-Francois, (1979),
“On Hotelling’s ‘Stability in Competition’,” Econometrica, vol. 47, pp.
1145-1150.
Economides, Nicholas (1984), The Principle of Minimum Differentiation Revisited,”
European Economic Review, vol. 24, pp. 345-368.
Economides, Nicholas (1996), “The Economics of Networks,” International Journal of
Industrial Organization vol. 14, no. 2, pp. 675-699,
http://www.stern.nyu.edu/networks/top.html .
Economides, Nicholas (1999), “Competition and Vertical Integration in the Computing
Industry,” in Competition, Innovation, and the Microsoft Monopoly: Antitrust in
the Digital Marketplace, Jeffrey A. Eisenach and Thomas M. Lenard (eds.),
Kluwer Academic Publishers 1999, http://www.stern.nyu.edu/networks/98-11.pdf
Economides (2001), “Industry Fragmentation After A Microsoft Breakup,” mimeo.
Economides, Nicholas, and Fredrick Flyer (1998), “Compatibility and Market Structure
for Network Goods,” Discussion Paper EC-98-02, Stern School of Business,
N.Y.U., http://www.stern.nyu.edu/networks/98-02.pdf .
36
Elzinga, Kenneth G., and David E. Mills (1999), “PC Software,” The Antitrust Bulletin,
vol. 44, page 739.
Evans, David S., and Richard L. Schmalensee, “Be Nice to Your Rivals: How the
Government is Selling an Antitrust Case Without Consumer Harm in United
States v. Microsoft,” in Did Microsoft Harm Consumers? Two Opposing Views,
American Enterprise Institute-Brookings Joint Center for Regulatory Studies,
Washington DC 2000.
Evans, David S., Albert Nichols, and Bernard Reddy, (1999), “The Rise And Fall Of
Leaders In Personal Computer Software,” mimeo., NERA.
Findings of Fact, United States v. Microsoft, Civil Action No. 98-1232 (TPJ) & State of
New York v. Microsoft Civil Action No. 98-1233 (TPJ).
Fisher, Franklin M., and Daniel L. Rubinfeld, “United States v. Microsoft: An Economic
Analysis,” in Did Microsoft Harm Consumers? Two Opposing Views, American
Enterprise Institute-Brookings Joint Center for Regulatory Studies, Washington
DC 2000.
Fisher, Franklin M., Written testimony, Civil Action No. 98-1232 (TPJ) & State of New
York v. Microsoft Civil Action No. 98-1233 (TPJ).
Hatch, Orrin G. (1999), “Antitrust in the Digital Age,” in Competition, Innovation, and
the Microsoft Monopoly: Antitrust in the Digital Marketplace, Jeffrey A. Eisenach
and Thomas M. Lenard (eds.), Kluwer Academic Publishers 1999.
Klein, Benjamin, (1999), “Microsoft’s Use Of Zero Price Bundling To Fight The
‘Browser Wars’,” in Competition, Innovation, and the Microsoft Monopoly:
Antitrust in the Digital Marketplace, Jeffrey A. Eisenach and Thomas M. Lenard
(eds.), Kluwer Academic Publishers 1999.
Lenard, Thomas M., (2000), “Creating Competition in the Market for Operating Systems:
A Structural Remedy for Microsoft,” mimeo. The Progress & Freedom
Foundation.
Levinson, Robert J., R. Craig Romaine, and Steven C. Salop, (1999), “The Flawed
Fragmentation Critique of Structural Remedies in the Microsoft Case,”
Georgetown University Law Center, Working Paper No. 204874.
Litan, Robert E., Roger G. Noll, William D. Nordhaus, and Frederic Scherer, “Remedies
Brief Of Amici Curiae In Civil Action No. 98-1232 (TPJ).”
McKenzie, Richard B. (2000), Trust on Trial, Perseus Publishing.
37
Reback, Gary L. et al., Memorandum Of Amici Curiae In Opposition To Proposed Final
Judgment, Civil Action No. 94-1564.
Plaintiffs’ Memorandum In Support Of Proposed Final Judgment, United States v.
Microsoft, Civil Action No. 98-1232 (TPJ).
Reddy, Bernard, David Evans and Albert Nichols (2000), “Why Does Microsoft Charge
So Little For Windows?,” mimeo.
Romaine, R. Craig and Steven C. Salop (1999), “Alternative Remedies for
Monopolization in the Microsoft Case,” mimeo.
Shaked, Avner and John Sutton, (1982), “Relaxing Price Competition Through Product
Differentiation,” Review of Economic Studies, vol. 49, pp. 3-14.
United States v. Microsoft, conference at New York University (May 5, 2000), in
streaming video at http://www.stern.nyu.edu/networks/video.html .