Is there such a thing as a software monopoly?

Is there such a thing as a software monopoly? By Mike Romano. Microsoft says no -- and its arguments could provoke changes in the antitrust laws.

By Mike Romano
Published November 11, 1998 8:00PM (EST)

Just before midnight on the eve of opening arguments in U.S. vs. Microsoft, Microsoft president Steve Ballmer e-mailed his staff to rally morale and outline the basic principles of the company's defense: "When you go home to your families, when you talk to your friends at the kids' soccer games or when you talk to business colleagues, remember how much people admire and respect our company," he wrote. "Microsoft's business practices [are] entirely consistent with the way other companies throughout our industry compete. There is no industry in America today that is more innovative and more competitive than ours."

Outside Microsoft, although the company surely has its fans, you have to hunt far and wide to find people who believe that it does not possess some kind of monopoly. But from where Microsoft sits, things look different: Microsoft has no monopoly because, as company spokesman Greg Shaw puts it, the very idea of a software monopoly is "a questionable concept. It's hard to even imagine a monopoly in software."

Microsoft's defense against the antitrust charges brought by the U.S. Department of Justice relies on the proposition that the software industry is competitive in the most absolute and technical sense of the word. As a result, Microsoft's defenders maintain, it enjoys unprecedented exemption from basic antitrust restrictions on practices like predatory pricing and bundling.

"Hopefully this won't sound like a metaphysical debate," begins Charles "Rick" Rule, who served as assistant attorney general in charge of the Justice Department's antitrust division under presidents Reagan and Bush and now represents Microsoft. First, Rule allows for the paradox that in such an inherently competitive industry it's still nearly impossible, or at least very difficult, to buy a PC without Windows. He concedes further that Microsoft intentionally undercut Netscape by giving away its Internet Explorer browser free. It's a textbook example of predatory pricing, he agrees -- if your textbook was written before the Internet. But Microsoft's lawyers maintain that, thanks to the unique nature of software as a product and the Net as a distribution system, the old rules no longer apply.

There is currently no statutory definition for predatory pricing. In 1993, however, the Supreme Court applied a standard test developed by professors Philip Areeda and Donald Turner in Brooke Group vs. Brown & Williamson, a case involving cut-rate generic cigarettes. According to Areeda-Turner, competition in any industry will naturally drive prices toward the marginal costs (i.e. the cost of material and labor in making the last widget, excluding the startup cost of building a widget factory). Pricing below marginal cost, at a loss, serves no purpose except to drive out competitors, and is therefore predatory. Once competitors are driven out of business, a predatory monopolist can recoup its losses by charging supra-competitive prices, to the eventual detriment of consumers.

Microsoft admits that it gives away Internet Explorer to cut into Netscape's market share and profit. (Netscape has charged up to $59 for its Navigator browser.) "But there's virtually no marginal cost in software," Rule points out. Making an extra copy of Internet Explorer costs practically nothing, and the Internet allows for free distribution. "What happened with Netscape was that Netscape initially gave away its software until it had a 70 to 80 percent market share, and raised its price," Rule explains in terms of Areeda-Turner. "Then Microsoft came along, providing competition, and drove the price to marginal costs -- zero."

By this reasoning, Rule argues that it's virtually impossible to price software below marginal cost -- and therefore, according to established definitions, predatory pricing is essentially nonsensical in the software business.

"I think Areeda-Turner is terrifically important," says NYU economics professor Lawrence J. White, who was chief economist for the Justice Department's antitrust division just before Rule took over and remains a harsh critic of Microsoft. "But it does create a dilemma for something like software," he says. "I rarely agree with Rick, but [predatory pricing] is not a useful way of thinking about this."

Indeed, the DOJ is not relying on predatory pricing charges in U.S. vs. Microsoft; instead, the government is trying a more nuanced case of illegal bundling -- tying an inferior product to a monopoly product. However, if you accept the argument that predatory pricing rules don't really apply to Microsoft, then the same circumstances responsible for that exemption -- free distribution and zero marginal costs -- also support the rather fantastic conclusion that monopoly power can't really exist in the software marketplace, either. And only monopolies are forbidden from bundling.

"It's dangerous to say never," says Rule, "but I don't think the notion of monopoly has a whole lot of weight in the current environment."

"Sure, most people look at the computer industry and say -- 'oh gee, Microsoft accounts for a very large amount of operating systems shipped on PCs today' -- but monopoly power is not defined by market share," Rule says, quite accurately. In antitrust law, monopoly power is the ability to set prices in a given market. No matter how little a competitor's market share, no matter that computer manufacturers testify that they have no practical alternative, Rule argues that because software distribution and marginal cost are negligible, competitors could instantly supply the entire market if Microsoft's prices became exorbitant. "Ergo, Microsoft doesn't have monopoly power to set prices," concludes Rule.

To prove its point, Microsoft's defense briefs -- and now the leaked "Halloween memos" as well -- repeatedly point to Linux in order to illustrate how cheap it can be to develop and distribute a fully functional operating system. (Don't forget that DOS cost Bill Gates only $100,000 in 1980.) Competition, even potential competition, Rule argues, abounds, restraining software prices across the board.

This sort of talk makes professor White apoplectic: "If you believe that, there's a bridge I can get you a reasonable price on. Everybody says it's easy to enter their market. It's bullshit." (Indeed, Coca-Cola once argued that neighborhood lemonade stands proved low barriers to entry in the soft-drink business.) White says that Microsoft's dominance doesn't come from its professed low prices or superior functionality, but from the intrinsic market inclination for software compatibility.

Demand for the first fax machine couldn't have been very high, White explains, but a network of fax machines exponentially upped the value of each individual unit. We don't buy Windows because we like it, the network theory goes, we buy Windows because everybody else has Windows. Developers build Windows applications because Windows has such a large installed base, and new computer buyers want a full array of applications -- so we buy Windows. Actual barriers to entry are therefore very high in the operating system market, says White, maybe impossibly high, after a standard is in place. Ergo, Microsoft is a monopoly.

"This is one proposition that anyone with a computer can evaluate as well as anyone else," says Kevin Arquit, former director of the bureau of competition for the Federal Trade Commission and current attorney for Sun Microsystems. "Frankly, I understand why Microsoft's saying it's not a monopoly. But by taking that point of view, it hurts their credibility across the board," he says. "They're just inviting some kind of change."

Such a "change," some observers suggest, could take the form of revisions in antitrust laws and regulations. "Attempts to apply conventional rules don't get you very far," explains UC-Berkeley economics professor Joseph Farrell, who testified at the Senate Judiciary Committee hearings on software competition in November. "I think there are legitimate questions for antitrust agencies to create some new guidelines or for Congress to possibly step in." A Judiciary Committee staffer who asked for anonymity said he didn't believe Microsoft's claim of exceptional status outside of conventional antitrust law stood "a snowball's shot in hell" -- but if the courts rule in favor of Microsoft's claim, "then yeah, we might look into something."

For his part, Judiciary Committee Chairman Orrin Hatch believes the century-old Sherman Act is flexible enough to rein in Microsoft. But he recently told a receptive audience at a Washington conference that, if the courts buy Microsoft's arguments, "in effect giving Microsoft a free pass to monopolize Internet-related technologies, [it] would have policy consequences of tremendous proportion, and would ultimately lead to some form of government regulation."

In other words, if Microsoft gets the courts to agree that there's no such thing as a software monopoly, the victory could provoke the other branches of government to rewrite the definition of monopoly for the digital age. And even if Microsoft wins, it could lose.

Mike Romano

Mike Romano writes about Microsoft for the New York Times, Wired magazine and Seattle Weekly.

MORE FROM Mike Romano

Related Topics ------------------------------------------

Department Of Justice Microsoft