Antitrust Law No Match for Microsoft’s Software Progress

It was 25 years ago that the Justice Department initiated its historic antitrust suit to “break up IBM,” accompanied by the huzzahs of Big Blue’s lagging rivals. When the suit was finally dropped as “without merit” in 1982, it had dragged on for 13 years and cost possibly a billion dollars. Lawyers had risen to senior partnerships in prestigious firms without ever working on another case. One MIT professor made so much money as an expert witness, he named his yacht after the antitrust statute.

Federal Judge Stanley Sporkin’s rejection of the antitrust settlement between the Justice Department and Microsoft Corp. last week conjures up the specter of another headlong tilt at competitive windmills. There are plenty of reasons to worry that Microsoft’s youthful founder and chairman, Bill Gates, will someday dominate the world software industry. But that does not mean he’s violated the antitrust statutes, or that an antitrust action is the way to put things right.

Back in the early days of the personal computer industry, IBM rather casually chose Gates’ then little company to produce the operating system for the IBM-PC. The operating system is the critical software that tells all a computer’s components what to do. “Application” programs, including word processors and spread sheets, won’t work unless they conform to the operating system’s rules. To IBM’s shock and lasting chagrin, Gates maneuvered his control of the operating system into a chokehold on the PC industry. For all practical purposes, almost every PC in the world, except those made by Apple Computer, uses a Microsoft operating system.

More ominously, Gates has leveraged his control of his newest operating system, called “Windows,” into a dominant share of the market for word processors, spread sheets and other applications. Presumably, since Microsoft is the expert on Windows, Microsoft software will work better with Windows than software from such companies as Lotus Development or WordPerfect. Based on sales of Windows-compatible applications, most customers accept that argument.

Gates’ competitors charge that Microsoft has competed unfairly--and they are right. Licensing practices, which Microsoft has already agreed to change, clearly placed some other companies at a disadvantage. There is also ample evidence that application developers at Microsoft know more about coming Windows versions than the competition did--which is also unfair. Finally, Microsoft is accused of “vaporware” announcements--announcing fictional new products to slow down a competitor’s sales--almost standard procedure in the software industry.

If the result of Sporkin’s action is that the Microsoft-Justice settlement will be tightened up a bit--particularly on the release of Windows information to competitors--no harm, and perhaps some good, will be done. (It’s hard to see how anyone can police vaporware announcements; telling the truth has never been a tradition in the software industry.)

But what competitors like Philippe Kahn, the erstwhile head of Borland Int’l, really want is to break up Microsoft--just like IBM’s competitors wanted to break up Big Blue. Specifically, they hope the government will force Microsoft’s operating-system business and the application-program business into two separate companies. That is a swamp the judge and the government’s lawyers, for the sake of their sanity, the public purse and the most efficient development of the software industry, had best stay away from.

A second issue before the Justice Department, but not covered by Sporkin’s action, is whether Microsoft should be allowed to purchase Intuit Inc., which has a dominant market share in personal financial software. There are anti-monopoly reasons why Justice might disallow that purchase. But the issues are separate from those before Sporkin.

The basic problem with a “break up Microsoft” strategy is that the computer industry changes so fast, and legal processes are so slow, that there is no way the two can ever be in sync. The distinction between an operating system and an application program is already dissolving. The newest operating systems from IBM and Apple, and the coming “Win95” release from Microsoft, include many features that used to be purchased separately as application programs, and this blending of operating and application software is accelerating. By the time a case is tried, in two years or so, its original premises may make no sense.

More important, the competitive landscape is in the process of rapid change. The “break up Microsoft” arguments seem to pretend there is still a battle for control of the desktop personal computer. That is long since over. Standard desktop applications are becoming commodities, profits are falling and companies like Borland are falling by the wayside.

The more interesting battles are taking place in the business market and the entertainment-video markets. Personal computing technology is displacing more traditional solutions in both markets, and Microsoft will be a formidable competitor. But it faces stiff competition on every level--operating systems, applications, networking software--and there is no assurance it will come out the winner.

Industrial-strength business software, for example, the kind banks use to run back offices, is far more demanding than anything Microsoft is used to selling, and it is having trouble adapting.

No industry in history has ever been subjected to the breakneck rate of change as the modern computer industry. Time and again, when a company seemed on the brink of winning total control of its industry segment--IBM, and earlier Apple, in personal computers, Lotus in spread sheets, DEC in mini-computers--it pitches into the mud the next day. The Justice Department and the courts have an obligation to see that Microsoft does not actually cheat. Policing the release of Windows development information, for example, will be in everyone’s interest.

But a healthy software industry will do the country more good than legions of antitrust lawyers. Gearing up for a long battle would be foolish in the extreme--or, in a favorite Gates-ism, “random.”