Remarks at the Cato Institution Policy Forum on Antitrust and Microsoft, April 20, 1998 (http://www.essential.org/antitrust/ms/catoapril20.html)
1. What is at stake?
In his day, John D. Rockefeller tried to monopolize oil production, refining and distribution. Alcoa sought to protect its monopoly in manufacturing aluminum. AT&T tried to monopolize local and long distance transmission of telephone calls and the manufacturing and sale of telephone handsets or devices that would connect to telephone lines. For decades IBM dominated the computer mainframe software and hardware market. Intel is trying to monopolize the manufacturing of hardware used to run personal computers.
Microsoft is more ambitious, and the implications of its global strategy are far more far important. Microsoft wants to use a core monopoly in software operating systems to dominate an enormous range of new and important areas of electronic commerce, and Microsoft wants to monopolize the software used to navigate the Internet and to navigate the next generation of television and multimedia programs.
If Microsoft were to succeed in every area it is active, it would have the most important control over commerce and worldwide information flows of any firm, ever.
Increasingly, we are talking about technology that is used as a gateway for many businesses, publishing ventures and civic communications, and Microsoft wants to dominate, influence or control the content itself, not just the transmission.
This elevates the disputes over Microsoft use of its Operating System as a matter of public policy.
2. How is this done?
In the software area, Microsoft engages in a very wide range of anticompetitive acts --- many of them are very similar to techniques used by Standard Oil, AT&T or IBM, before each of these companies faced antitrust action.
One strategy of a monopolist is to deter entry or investment by rivals by engaging in predatory pricing. Standard Oil used cross subsidies to selectively cut prices, so it could bankrupt its rivals. AT&T used cross subsidies to selectively cut prices, and drive rivals out of business. IBM used cross subsidies to selectively cut prices, and drive rivals out of business. Microsoft does this too.
Microsoft can take a rival's core product, and spend countless millions in R&D or acquisitions, and then offer a competing product for free, or bundle it with Windows or with Microsoft Office --- the suite of office productivity applications which are nearly as ubiquitous as the operating system.
For example, faced with Microsoft's decision to spend hundreds of millions of dollars on a free alternative, Netscape is unlikely to justify continued R&D spending on its browser. And when Microsoft announced that it would include copies of its Outlook product in Microsoft Office, Netmanage announced it would discontinue further development of Ecco Pro (http://www.netmanage.com/products/eccopro/).
Microsoft benefits from predatory pricing in two ways:
In high tech markets, it is often the case that products must interoperate with each other. AT&T tried to limit the ability of competitors products to interoperate with the AT&T telephone network, by withholding technical information, using proprietary technologies, or by changing standards to create incompatibilities of rivals products. IBM did this. Intel is doing this now. Microsoft has done this for a long time, going back to the days when programmers coined the phrase, "DOS isn't done until Lotus won't run," referring to Microsoft's introduction of minor changes in DOS that created problems with Lotus 123, the spreadsheet program that is a competitor to Microsoft's Excel. If Microsoft can't make its own products look better by taking advantage of technical back doors, it can make a rivals products perform badly, or it can make its own products technically essential, as it is trying to do with the browser.
When you combine both predatory pricing and technological predation, firms and investors decide on their own to keep out of Microsoft's way. Consider the following quote from an April 18 story by Lisa Bowen in Ziff Davis's ZDNET web site: (http://www.zdnet.com/zdnn/content/zdnn/0418/308142.html)
A further challenge for the DOJ is showing that Microsoft is actually stifling innovation because it's hard to measure Microsoft's products against those that never make it to market.
There's no question that many developers are shying away from independent projects in areas that Microsoft might consider. At Microsoft's Windows CE development conference, developers lined up during a question-and-answer session to ask the software behemoth which products it doesn't plan to develop, as if they were looking for crumbs.
Other software makers said they attended the conference to check out Microsoft's plans to make sure they stay out of the company's road.
A new area of predation for Microsoft concerns Internet navigation. One of the major reasons that Microsoft wants to have a monopoly on Internet Browsers, is so that Microsoft can design the Windows operating system to have as much control as possible on navigation itself. There are several aspects of this.
Microsoft wants to write the default bookmarks and menu options for content based upon current and new Internet technologies. The unsuccessful experiments with so called "push" channel technologies was one attempt. The new Microsoft "Start" page project is a more elaborate version. Microsoft has also designed its Browser so it can periodically check in with Microsoft to reset bookmarks, menus and other items with ones Microsoft's suggests, gently but ever so steadily taking consumers "where Microsoft wants them to go today." Microsoft is developing its own search engine, which it hopes will replace Yahoo and other popular search sites.
This is a "path of least resistance" strategy, based upon the idea that time and attention are the ultimate scarce resource in the information age.
Will it work? Consider commercial airline reservation systems. One study indicated that professional air travel agents using online reservation systems would pick the first fare they saw 53 percent of the time, and a fare from the 1st screen 93 percent of the time.
When Microsoft bought Web TV, it changed the travel menu so that Expedia, the Microsoft travel service, appeared first in the travel menu. Sabre told us their Travelocity web site lost is prominent menu location, and was moved to page 6, next to Tom's Travel, in an alphabetical listing. How many people in the audience ever look at page 6 when you use an Internet search engine?
What will happen if Microsoft succeeds in its wildest dreams and determines which flower shop, which citizen group, and which car dealer appears on page 1 and which one appears on page 6? What if Microsoft could determine what information appears on page 1 when a person searches for information about Representative Rick White or legislation concerning digital copyright?
3. What Should Be Done.
In 1997, we organized a conference to Appraise Microsoft's Global Strategy. I believe we now have a fairly good idea of where some of the problems are. It is time to shift the debate to the issue of remedies. What can and what should be done about the Microsoft Monopoly? This will be the focus of our next Microsoft conference.
The current DOJ litigation deals with narrow issues concerning restrictive contracts and product bundling. The easiest remedies would limit the use of restrictive contracts, such as contracts that prevent Internet Service Providers or OEMs from giving consumers the opportunity to choose non-Microsoft products. Practical rules regarding product bundling are more difficult, as is the issue of predatory pricing, which DOJ and the EC have ignored. But there are several other types of remedies which may be more useful.
Issues regarding interoperability are very important. The European Community's 1984 undertaking with IBM was mostly about interoperability issues. (http://www.essential.org/antitrust/ms/1984ibmeu.html). There has been much antitrust work on interoperability that relates to telephone monopolies. In recent years, the Federal Trade Commission has negotiated agreements with several other software companies to open user interfaces, such as the FTC's 1995 agreement with Silicon Graphics, Inc (Docket No. C-3626), which required SGI to "establish and maintain an open architecture, and publish Application Program Interfaces ("APIs"), for . . . computers and operating systems in such manner that [third party] software developers and producers may develop and sell . . . software, for use on [SGIs] computers, in competition with [SGI]."
More relevant is last week's landmark ruling in the Intergraph v. Intel case, where a federal judge ruled that Intel's CPU platform is an essential facility - and ordered Intel to provide non-discriminatory access to technical data needed to develop products which interoperate with the Intel CPU. (http://www.intergraph.com/intel/highlights.stm ). Now that the "tel" half of Wintel is considered an essential facility, what about the "Win" half? Biases of Internet navigation and related areas are very important, particularly if Microsoft succeeds in monopolizing the browser market, dominating the search engine market, and becoming the front end for new video set top boxes. Policy makers and the public need to debate conduct rules which would prevent a dominant OS vendor, like Microsoft, from exercising undue influence over Internet information searching and navigation technologies.
There are also other remedies that challenge Microsoft on a more basic level. We are asking OEMs to offer consumers the opportunity to purchase alternative operating systems, not owned by Microsoft. (http://www.essential.org/antitrust/ms/ipnmarch91998.html) These include both commercial competitors, like Rhapsody, BeOS or OS2, and a new generation of powerful free operating systems, such as Linux or FreeBSD, which are rapidity maturing as alternatives.
We believe there are factors which make it more feasible for a new OS to succeed. Larger and cheaper hard disks and computer memory make it possible to run multiple operating systems on the same computer. We do this now at our offices. Secondly, the Internet and new Internet standards bodies make it easier to share data across OS platforms. Third, new software development tools make it easier to port software applications across platforms.
There remain barriers for new OS platforms, however. The most important of which concern device drivers, which are still scarce for Windows alternatives. Making matters more difficult would be efforts by Microsoft and Intel to control the architecture of a new generation of high performance device drivers.
Leading Original Equipment Manufactures (OEMs) for personal computers, like Dell, Micron, Gateway 2000, Packard Bell, Compaq and others need to be free from retaliation by Microsoft if the OEMs offer non-Microsoft products, including choices of software operating systems. This may be difficult in practice if Microsoft can discriminate in its pricing of the software OEMs need for the current corporate and consumer market. If Microsoft were required to use non-discriminatory licensing of Windows and Microsoft Office, OEMs would be free to offer consumers additional choices.
Essential Information has created an Internet email list to discuss these issues. You can participate by sending a note to email@example.com with the message "sub am-info yourfirstname yourlastname." Archives of this list are available on the Internet at http://www.essential.org/listproc/am-info/ (no period). The Consumer Project on Technology also maintains a web page on Microsoft antitrust issues at http://www.essential.org/antitrust/microsoft/microsoft.html (no period).