A Framework for Negotiation on a Microsoft Remedy
Hal R. Varian
December 1999
If Microsoft is found guilty of violating Federal and/or State
antitrust laws, the remedy will certainly involve conduct
requirements. For example, transparent pricing and contracts (``open
Windows'') and full documentation of Windows APIs and file formats
will likely be required.
But the Department of Justice will probably want to ask for more than
this. The difficulty that they face is that other proposed remedies
such as source code licensing or breakup are very big steps indeed,
and they run the serious risk of standards fragmentation.
One remedy that is less dramatic than breakup or source code licensing
that has not been widely discussed is something called the
``compulsory licensing of old binaries remedy,'' or CLOBR. An example
would be the following: ``Microsoft is required to license binary
versions of its operating system more than 36 months old at a flat
price of $10 per license.'' The 36 months and the $10 figures are
only meant to be illustrative; they could easily be adjusted.
CLOBR has several interesting features:
- It creates competition for current versions of Windows without
creating significant standards fragmentation.
- It discourages Microsoft from unnecessarily bundling additional
components into Windows, since they will have to sell the bundled
package 36 months later for a low price.
- CLOBR leaves Microsoft ``free to innovate.'' In fact, it virtually
forces Microsoft to innovate since they will need to ensure that each
new version of the OS is substantially better than the widely
available older but cheaper versions.
- It leaves Microsoft free to compete aggressively in other markets,
while preventing it from using its dominance of the desktop market
to gain unfair advantage over competitors, since an inexpensive
alternative to the current version of the OS is always available.
- It is highly tunable, unlike other proposed remedies. A
requirement to license 5-year old binaries at a price of $40 would
have no impact on Microsoft at all, while a requirement to license
2-year old binaries at $5 would be quite significant. This makes
CLOBR a useful framework in which to discuss a negotiated settlement.
The Court could even reserve the right to modify the terms later on,
based on Microsoft's subsequent conduct.
But CLOBR, like other proposed remedies, has faults.
- The presence of the older, inexpensive OS would create competition
for Microsoft, but also create competition for other new entrants into
the OS market.
- Under CLOBR, Microsoft would have an incentive to unbundle
features from the core OS and sell them as add-ons. This incentive
would have to be countered by a definition of what constitutes a
minimal operating system. Note that unlike the browser debate, the
issue would revolve around a definition of a minimal OS, rather than
the maximal OS.
CLOBR's logic is similar to that of patent or copyright law. In
intellectual property law, an inventor is granted a monopoly for a
fixed period of time to its intellectual property so as to provide
appropriate incentives to innovate. But the monopoly is limited, so
as to avoid the social losses from monopolization. In the case of
patents and copyright, the innovation eventually passes into the
public domain. In this case, the software passes into the compulsory
licensing phase, rather than the public domain, but the effect is
similar: the temporary monopoly in the intellectual property provides
a balancing of private incentives and the public good.