Microsoft Is Bad, Uncertainty Is Worse

Chris Anderson
Wired

August 2002

Money, ambition, caffeine: All sorts of forces drive Silicon Valley. Not least among them is the dark force - hatred of Microsoft. Software engineers can't agree on much, yet in geek circles it's an article of faith that Microsoft's products are mediocre and its tactics mean. Only the brave dare to combine the words "innovation" and "Redmond" with anything but sarcastic intent.

This is why the US prosecution of Microsoft for monopolistic abuse was so widely cheered in the Valley, and why its whimpering end is seen as such an anticlimax. There are plenty of other furies out there — especially holdout states and the European Union — who will continue to press their cases, perhaps for years to come, but the Feds have had enough. With each step toward a weak resolution, the gloom darkens in the Valley: The bad guys will win in the end. Sigh.

There is, however, a more pragmatic way to look at the case: that it had reached the point of diminishing returns. From this view, the sheer length of the process threatened to overwhelm the benefits derived from its outcome, and the sooner it ends, the better, regardless of how imperfectly. Justice is ideal; however, antitrust is about economics. Much of the long debate has been about whether the market would benefit from breaking Microsoft in two. That's a commendable question, and the very act of asking it has a cost of its own. Each year the case has dragged on - the current marathon started in 1997, and the first probe in 1990 - has been another year of uncertainty for the technology industry. And uncertainty can be expensive, now more than ever.

The networked economy is most powerful when absorbing information at light speed. No news is another matter. In a system tuned for just-in-time manufacturing and millisecond trades, not knowing is costly.

As the case drags on and on, everyone suffers

The effect is hard to quantify. It's the dog that doesn't bark: funding not offered, businesses not entered, innovations not pursued. Antitrust is meant to help the overall market, for the ultimate good of the consumer. But the fast-moving technology industry is hypersensitive to delay. A legal process that's right for steel is too slow for software.

How slow is too slow? Economists George Bittlingmayer of the University of Kansas and Tom Hazlett of the Manhattan Institute looked for answers in the Nasdaq. Their research, a version of which appears in the new book Microsoft, Antitrust, and the New Economy, is interesting, though far from conclusive. If shareholders believe that the antitrust process is likely to help Microsoft's competitors, as intended, shares in those companies should rise when the case turns against Microsoft. The opposite has turned out to be true.

"Pro-enforcement actions" hurt Microsoft, as one might expect; they also hurt its rivals. "Anti-enforcement actions" helped both. The same effect played out over the entire Nasdaq. From March 1998 to April 2000, the Nasdaq (excluding Microsoft) lost $57 billion on days when the tide turned toward the prosecution and gained $623 billion when it went the other way. The numbers may not capture the whole picture (the Nasdaq was in a record boom), says Bittlingmayer, yet there's clearly something going on here.

His explanation: The market had little confidence in the efficacy of antitrust, at least in this case. Government victories were seen as continued uncertainty; Microsoft's wins were hastened resolution. Of course, Redmond shares the blame. It has gummed up the process with appeals. But it's in the nature of a company to defend itself with every weapon it has.

So who was looking out for the economy as a whole? The Justice Department, the states, and the EU all aim to put consumer interests first by ensuring sufficient competition. Instead, the combination has generated a costly mess. The states don't agree with the Feds, and the EU marches to its own beat on such big cases (witness last year's Honeywell fiasco).

Doctors have a Hippocratic oath, swearing first to do no harm. Regulators do not, and we all pay the price. Trustbusters have calculated the costs and benefits of breaking up Microsoft or, for that matter, letting it run free. Few have looked at the cost of doing neither for a long, long time. Whatever you may think of Windows, this is worse.

Chris Anderson (canderson@wiredmag.com) is Wired's editor in chief.