Antitrust in the Information Age


Despite intense competition and absent any evidence of consumer harm, the Justice Department persists in its antitrust litigation against Microsoft. Today, I propose to address three questions about that litigation. First, what theory does the government rely on to defend its interventionist role? Second, what logic, if any, justifies the remedies reportedly under consideration by the Justice Department? Third, what's really driving the browser wars?

I turn first to the theory of network effects, which says that high-techconsumers, concerned about compatible software applications, might beseduced into purchasing inferior goods. Undoubtedly, there's some truth tothe theory. But in unregulated software markets, users were not inveigled bynetwork effects despite compatibility concerns. Consumers refused to anointWordPerfect as ruler-in-perpetuity of word processing, or Lotus asspreadsheet leader forever, merely because usage of those products wasubiquitous.

Network effects theorists argue that consumer purchases are not a reliableindicator of a "superior" product. Instead, objective measures can tell uswhich products are technically superior, no matter what consumers actuallybuy. The problem with that argument is that it leads directly to governmentpaternalism -- to the idea than an elite corps of government experts knowsour interests better than we, and can regulate our affairs to satisfy thoseinterests better than the market does. When we permit government to makesuch assessments, and we allow those assessments to trump the subjectivechoices of consumers, we abandon any pretense of a free market.

Next, let me address the various remedies that have been floated by theDepartment of Justice. Essentially, each remedy punishes vigorouscompetition by dismembering the winning competitor. First, verticaldivestiture: One company keeps the Windows operating system; a secondcompany gets applications programs; a third company takes on Internet ande-commerce products. Normally one doesn't attack monopoly power by spinningoff the monopoly itself into a separate company. But what is worse, verticaldivestiture will require ongoing government decisions about whether aproduct is part of the operating system, or an application or Internetrelated. Just look at the browser to see how difficult it can be tocompartmentalize a product within a nearly seamless operating environment.And look at the AT&T fiasco to see how easy it is for a court to get boggeddown in post-divestiture regulation.

Second, horizontal divestiture: Each of several vertically integrated clones-- Baby Bills -- would receive full rights to Microsoft's source code andother intellectual property. They could then proceed to compete freely andfiercely against one another. May the better Bill win, at least until a newleader emerges, at which time the government will undoubtedly call foranother divestiture. No one seems to know which corporate Bill gets thereal-life Bill, nor whether new operating system features have to be sharedand, if so, why any company would continue to innovate, knowing thatcompetitors will reap the benefits.

Third, compulsory licensing: Microsoft would be forced to license theWindows source code to one or more companies, each of which could developand sell it independently, thereby creating instant competition in theoperating system business. The problems should be obvious: If new technologyis to be declared public property, it will not materialize. If technology isto be proprietary, then it must not be expropriated. Once expropriationbecomes the remedy of choice, the goose is unlikely to continue layinggolden eggs.

Meanwhile, government-driven balkanization of operating system protocolswill wipe out Microsoft's most important contribution to software markets:standardization. Like the Unix system, Windows will end up with a dozen ormore variations -- no common platform on which software developers canbuild. The result will be fewer applications programs, increased costs ofdevelopment and higher user prices.

Finally, what's really driving the browser wars? To find out, just read theJustice Department's complaint. There you will find Netscape mentioned 130times in 130 pages -- government resources co-opted for the welfare of acompetitor, not for consumers. The conclusion is all but inescapable thatantitrust laws are being used as an anti-competitive subsidy to prop upunsuccessful firms. Instead of focusing on new and better products, softwareexecutives will find themselves consorting with former members of Congress,their staffers, antitrust officials and the best lobbying and publicrelations firms that money can buy.

On the other hand, perhaps this case is merely support for empire building,which has become standard procedure in Washington, D.C. President Clintonhas asked for a 17 percent increase in funding for the Antitrust Division.To justify that increase, the Department of Justice must provide theAmerican public with dramatic evidence of its effectiveness. Hence, ahigh-profile case with sensational remedies as the exit strategy, played tothe media and focused not on substantive legal issues but on public ridiculeof a company and its chief executive. We deserve better.

These remarks were given by Robert Levy on May 24, 1999, in the United States Capitol at the National Taxpayers Union Conference titled Antitrust in the Information Age.