‘Baby Softs’? Imagining a Breakup of Microsoft

May 25, 1998

Now that the Justice Department and the attorneys general of 20 states have filed long-threatened antitrust suits against Microsoft Corp., the question remains why the most obvious, and perhaps cleanest, solution has yet to be put on the bargaining table: the breakup of Microsoft into its component businesses.

Many analysts and industry executives have noted that the complex, arcane
lawsuits filed last week are probably too closely tied to the software giant’s
current products and business strategies to have any lasting effect. Demands
regarding Windows 98, for example, or even specific concessions to
Microsoft competitors like Netscape Communications Corp., will probably be
outdated in a matter of months as technology continues to change at its usual
breakneck pace.

But perhaps most important, the challenges to Microsoft’s monopoly do not
address the more fundamental issue of how to insure a truly competitive
environment in which software code is the currency upon which a new digital
economy is increasingly based.

Microsoft’s iron control of the market for operating-system software is the
keystone of its continuing efforts to dominate applications programs, new
media and other businesses — a power that has prompted speculation about a
breakup for years.

Most recently, a Silicon Valley research institute, SRI International, offered a
hypothetical model in which Microsoft eventually becomes five separately
owned and operated companies, which SRI called the Baby Softs.

According to SRI, these Baby Softs would consist of the following:

— An operating-systems company that sells all versions of Microsoft’s
Windows operating system.

— An applications-software company that invents, acquires and sells classical
productivity applications like word processors, spreadsheets, data bases, and
Internet browsers.

— A content media company, taking as its genesis today’s Microsoft
Network, that would acquire, promote and sell access to online and
traditional media products.

— A networking company that includes network technology assets like the
recently acquired WebTV Networks.

— A company that oversees the holdings of Microsoft’s chairman, Bill Gates,
and anything else that does not fit into the other companies.

The logic of such a setup becomes apparent when viewed against the
specific demands of the lawsuits filed last week. Most of the demands in the
suits are intended to stop Microsoft from using its operating-system
monopoly to force personal computer makers and consumers to accept the
company’s applications and World Wide Web links.

To preserve competition for Internet browser software, Justice Department
officials have asked that Microsoft either decouple its Internet Explorer
browser from Windows 98 or include Netscape’s browser. Although this
demand might provide an unnecessary entitlement for Netscape, it still does
not provide for what happens when Microsoft ships future versions of the
operating system like Windows NT or Windows 2000 — which, like
Windows 98, will incorporate Microsoft’s browser in such a way that it
appears to be an integral part of the operating system.

The Justice Department has various other goals, including forcing Microsoft
to let computer makers distribute rival browsers if they so choose
— something the company has contractually forbidden manufacturers to
do in the past. The states, in their lawsuit, want to stop Microsoft
from linking the sales of its Microsoft Office application suite to sales
of Windows for computer makers.

Breaking up Microsoft into separate operating-systems and applications
companies, with a clear definition of those two categories, would eliminate
the need for all these demands. Leaving Microsoft intact fails to address the
original impetus for the antitrust suits: the need to create an environment
in which competition can flourish.

In fact, in seeking to re-establish a competitive market, the definition of
an operating system versus an application is a crucial component of any legal
action against Microsoft, and perhaps the most enigmatic.

This is a contentious area, and one in which Microsoft has to date been
intractable. It is none of the government’s business, the company says,
to decide what it will build into its operating-system software, including
applications like browsers.

But the malleable nature of software is such that when lines between
operating systems and applications blur — as they have so fully to date — the
ability of Microsoft’s competitors to provide innovative applications, and the
ability of consumers to buy those products, is all but eliminated.

The argument that an operating system is whatever Microsoft or anyone else
says it is has purely technical merit. But one can imagine that AT&T Corp.
felt the same way about what constituted a telephone network in 1974, when
the government first filed suit against the telephone monopoly for keeping
competitors out of the now fiercely competitive long-distance and
telephone-equipment markets.

In the end, defining the components, and the boundaries, of the software
industry will be necessary to level the playing field for Microsoft’s
competitors and to create a foundation for the future of digital commerce.
Rather than cutting incremental deals with Microsoft about specific products
and practices, the government should ask: In what areas are there companies
that want to compete? Where are competitors being unfairly stopped by an
existing monopoly — or where might that monopoly unfairly stop them in the
future — and where and how do we draw the line?

“The worst thing would be a settlement agreement with Microsoft,” said a
prominent Silicon Valley lawyer who contends that the federal and state
governments have not gone far enough in their lawsuits. “There needs to be a
bright line drawn that can’t be interpreted through wordsmithing,” he said,
speaking on condition of anonymity.

“If there are ambiguities on a piece of paper, Microsoft will just do what they
want and argue later about the paper. The government needs to say: ‘These
are the places where you cannot go. Don’t go there!”‘

Under that reasoning, a true challenge to Microsoft’s monopoly would require
that the government set new definitions and standards not only for operating
systems but also for all the product lines and business strategies by which
Microsoft and its competitors conduct themselves in the future.

Certainly, it would be no picnic for the government to enter a definitional
debate with Microsoft in an attempt to draw that bright line separating its
operating system from other software products.

But by not addressing these fundamental issues, the government could
squander a crucial opportunity to build a firm and lasting foundation for fair
competition in the digital age.

Denise Caruso

Copyright 1998 The New York Times Company