Has Google actually read U.S. v. Microsoft?

Policy analyst James V. DeLong says Google's real game is to blur boundaries between desktop search and Internet search.
Written by James V. DeLong, Contributor

perspective In the latest clash of titans, Google has filed a brief with the federal court overseeing the consent order in the Microsoft antitrust case.

It argues that enforcement of the consent order should be toughened, and its duration extended beyond the current November expiration date. Because several states--including Google's home of California (ah, coincidence is wonderful)--endorsed the idea of an extension at the status hearing on September 11, and the judge is deliberating, the attempt must be taken seriously.

Google's objection concerns desktop search. Microsoft's latest operating system, Vista, has such functionality and, of course, Google is in the business of distributing a desktop search program, with particular attention to linking it with Google's global Web searches.

Google cannot complain that its desktop search does not run on Vista; it runs just fine. Nor can Google protest that its offering is somehow blocked; Vista permits end users and original equipment manufacturers to select an alternative desktop search program. Nor can Google claim to have been blindsided by the issue; Microsoft worked with Google and other companies along with various governments, on the search issue. For its part, the U.S. Department of Justice, along with most of the states, appears satisfied with the resolution.

So what is really going on in Google's collective mind? I think it boils down to a couple of things. First, Google and Microsoft are colliding as the tech world evolves, so disrupting Microsoft's plans is its own reward. Second, Google has a dominant position in general Web search. Integrating Web search closely with desktop search would reinforce and extend this power. An effective way to accomplish this is to force Microsoft to do it in the name of "fostering competition."

The weaknesses in Google's argument are that it has nothing to do with the original Microsoft case or with the purpose of the consent order, and that it does not relate to the conditions of the market that have developed in the years since the case ended.

The litigation determined that Microsoft had a monopoly over operating systems for Intel-based PCs. But having a monopoly is no offense under the antitrust laws when it was honestly attained by superior skill and foresight, and this was clearly true of Microsoft. Historian Martin Campbell-Kelly notes that as of 1985, two dozen alternatives to MS-DOS were offered by more than 20 vendors, many of them bigger than Microsoft. Microsoft won by being the best at meeting the needs of consumers and developers.

The only antitrust charge against Microsoft that survived was that it had acted to maintain its monopoly. In particular, it hobbled the use of "middleware," which was defined as software that could serve as a bridge between applications and the Windows operating system. In accord with the crucial role played by Netscape and Sun Microsystems in persuading the Justice Department to bring the case, Navigator and Java were singled out as examples of middleware.

The government's theory was a sort of double bank shot--if enough application writers chose to link to middleware rather than to Windows itself, then the middleware programs could take all these apps with them and decamp to link with some other operating system, which would then provide competition to Windows.

This theory has some problems. For example, it would transfer any monopoly power to the middleware, and why should either apps or operating system developers or the government want that? But it was close enough for government work, and became a key part of the consent order. Microsoft is under strictures to ensure that middleware works with Vista.

The goal of the decree remains: to allow the proliferation of middleware, which will in turn provide a mechanism that will allow the development of competitive operating systems for Intel-based PCs, if anyone is inclined to create such systems.

Middleware has indeed flourished during the past six years, whether as a result of the order, the inherent dynamism of the tech industry, Microsoft's own need to nurture developers, or all three. Each new PC contains an average of 35 pre-installed non-Microsoft programs, and in four out of five middleware categories non-Microsoft products outstrip those from Microsoft.

In reality, the very existence of Google renders most of the Microsoft consent order unnecessary.

This eruption of middleware has not produced any spate of new operating systems, and there is serious doubt that the theory of the double bank shot was a sound one.

But, as noted, it was never a goal of the consent order to reduce Microsoft's market share. The goal was to allow the market to work freely, and that has been achieved.

Even if one takes a broader view that the point was to create checks on Microsoft, that is certainly occurring. The iPod has given Apple new pizzazz. Judging by the commentary about Linux coming from the free software community, Microsoft is already dead. Another new big thing is "software as a service," which could reduce a desktop operating system to a few lines of code to connect peripherals. The price-earnings ratio for Microsoft stock is less than 20, so Mr. Stock Market doesn't see the company as possessing much pricing power.

Perhaps the most important development is Google itself. In 2001, Microsoft argued that the existence of middleware companies with links to the applications developers contradicted the conclusion that it had monopoly power. The Court of Appeals refuted this by noting that middleware companies that provided an alternative to Microsoft's own middleware offerings did not actually exist, and that one of Microsoft's sins lay in "keeping rival browsers from gaining the critical mass of users necessary to attract developer attention away from Windows as the platform for software development."

Now, not only has middleware grown, but Google's desktop search program, whether or not technically "middleware," provides a platform to which applications could attach, and, as Google says proudly in its intervention brief, it already has critical mass: "Hundreds of millions of...copies have been distributed to users," and "already, third-party developers have written thousands of applications using APIs in Google Desktop."

In reality, the very existence of Google renders most of the Microsoft consent order unnecessary. If Google were interested in promoting a competitive desktop operating system, it could create one or it could throw its weight behind Linux. In either case, its weight includes power over Internet search. Mr. Market agrees on the relative power of the two companies, because it gives Google a price-earnings ratio of roughly 46, more than double that of Microsoft.

So what is really going on, if Google is not interested in operating systems (much better money in search)? The real game is to blur all boundaries between desktop search and Internet search, so as to use Google's existing market power to maximum leverage. If this twists the Microsoft consent order away from its proper focus, and turns it into a device for suppressing competition in search by preventing Microsoft from developing a superior desktop product, which it then expands into search, that is a bonus.

An ironist might call this monopoly maintenance by Google. Perhaps antitrust fans can anticipate a U.S. v. Google, in which Exhibit 1 will be the intervention brief.

James V. DeLong is special counsel for Kamlet Shepherd & Reichert. He is also vice president and senior analyst for the Convergence Law Institute. The opinions expressed here are the author's alone.

This article first appeared on ZDNet Asia sister site CNET News.com.

Editorial standards