Is One Microsoft Enough?
       click here to go back to the home page

also Here's a modest proposal for helping the world cope with Microsoft. Split the company into triplets—three equal, cloned Microsofts. Robert Bork, once upon a time too conservative to get onto the Supreme Court, has been floating this, speculatively, he says, since the company hasn't yet been found guilty.
   It sounds scary, at first: three new companies, each starting fresh with its own copy of the Windows source code, and each with its own copy of the gene for Determination to Vanquish All Competitors and Control the World. There would be complications, for sure. The three new Microsofts couldn't really be identical. For example, only one of them (call it "Huey") could have Bill Gates himself, the chief executive once famous for holding a tactical and strategic roadmap of the entire computer industry inside his head. Dewey and Louie would have to go scrounging for CEO's of their own. Then again, maybe that's not so important; after all, in his testimony for the ongoing antitrust case, Gates has sworn that he didn't actually know much about his company's key business decisions. ("Q: Did you ever try to find out? A: I read something that was on our Web site about four days ago"; "Q: Are you aware of an agreement that Intuit entered into with Microsoft? A: I know there was some kind of an agreement. I wasn't part of negotiating it, nor do I know what was in it.")
   Maybe one of the companies could get the kind of chief executive who would be willing to show up in court, take responsibility for the business, and explain its actions forthrightly and enthusiastically and proudly.

Meanwhile, to avoid colluding with himself, Gates might have to sell his stock in all but one of the new Microsofts—at an enormous profit. "But that's not an objection," notes Bork. "If Bill Gates gets even richer, that doesn't violate antitrust laws."
   Competing software companies might not rejoice at the prospect of confronting three fierce and powerful opponents where now they have just one, but that doesn't trouble Bork, either, though he is working as a paid consultant for Netscape. "Preserving Netscape is not the object of the antitrust laws," he says. "The object is consumer welfare."
   Earlier proposals for breaking up Microsoft have tended to imagine separating the different pieces of its well integrated empire: one company for the operating system, and another for applications, and maybe another for Internet services. That has a kind of natural logic, because Microsoft is accused of leveraging its control of one realm to gain advantage in the others. But this approach—the so-called "vertical" split—might be messy and hard to enforce. Microsoft has taken pains to stress how difficult it is to draw boundaries; yesterday's application tends to become today's operating-system feature.
   A horizontal split—the cloning proposal—would let the Government step aside and stay out of the software business. Set the three—or five, or seven—new Microsofts loose and let them compete, with one another and with everyone else. That's what happened in 1911, when the early trust-busters succeeded in breaking up Standard Oil, which dominated the oil business almost as comprehensively as Microsoft does personal-computer software.
  One beneficiary would be the PC manufacturers. They would suddenly have a choice of suppliers of the operating system without which their machines don't run. Their unaccustomed new bargaining power would presumably push prices down and also let them refuse some of the onerous contract restrictions that Microsoft has been able to impose. Microsoft maintains that Windows is already a great bargain, but it refuses to disclose how much money it makes from the operating system; Wall Street analysts know only that the company as a whole has exceeded their earnings forecasts for 27 consecutive quarters (good thing these analysts aren't trying to make a living handicapping football games).
   Maybe the new companies would compete by specializing. Certainly any new versions of Windows or of Microsoft's ubiquitous cash-cow Office suite would have to be compatible with the versions everyone has now—any drastic change would be suicide. But Huey might decide to be known for quality and stability: instead of adding bulky new features, it might create a lean and crash-proof version. Dewey could favor futuristic experimentation and power users. (Maybe Louie could specialize in the lyric poetry of error messages; it would be a shame if the world forever lost this specimen of Microsoft wisdom, from Word: Permanently changes the selected Unknown into the object type Unknown.)

In the long run, Microsoft's shareholders might thank the Antitrust Division. Maybe even in the short run—it's hard to imagine the stock market betting against the three clones.
   Microsoft has said many times that it already lives in a highly competitive environment; that it wants nothing more than the freedom to "innovate" on behalf of consumers; that it seeks no particular advantage from its monolithic dominance over software markets. A horizontal breakup would create the world the company likes to pretend already exists. Each new Microsoft would start life with an enormous stockpile of cash, an impressive reservoir of talent, rights to what may be the industry's most formidable portfolio of software patents and, most important, a genuine need to innovate on behalf of consumers.
   They just wouldn't wield quite so strong a hammer. Huey, Dewey and Louie could theoretically engage in actions that might be illegal for a monopolist. They could change their operating system to sabotage audio and video players from companies like Apple and Real Networks, as Microsoft did. They could attempt to bludgeon chip makers like Intel with threats to withhold support for their microprocessors. They could selectively withhold technical information or demand control of PC manufacturers' start-up screens. They could do all these things legally, because they would no longer be a monopoly; but they might not try—because they would no longer be a monopoly, and so, for the first time in years, all these companies would have an opportunity to say no.
   "I think it would be good for everyone but Microsoft," says James Love, director of the Consumer Project on Technology. "Consumers might choose the company that was least predatory. It would be easy to boycott the 'bad Microsoft' if there were at least one company that wasn't so bad."

Copyright 1999 James Gleick
First published in the New York Times Magazine 14 February 1999