Does A Breakup Make Sense?

Taming Microsoft may require the ultimate penalty. The trick will be to also promote innovation

Everyone knew that the government had mounted a strong case in U.S. v. Microsoft. But nobody expected a complete rout.

Yet, when U.S. District Judge Thomas P. Jackson released his finding of facts on Nov. 5, he handed state and federal prosecutors an unambiguous triumph. Declaring that Microsoft routinely used its monopoly power to crush competitors, he portrayed the software giant as nothing less than a social menace.

Jackson's fact-findings were so critical of the company that they've raised the stakes in this battle: Suddenly, the breakup of Microsoft is a real possibility. "If you had asked me how likely a breakup of Microsoft was six months ago, I would have said 10%," says Robert E. Litan, an antitrust expert at the Brookings Institution in Washington, D.C. "Now I say it's 50%." Even Justice antitrust chief Joel I. Klein confided to millions of Sunday morning talk show viewers that a breakup is now squarely "in the range" of potential solutions.

While it's clear that the idea of a breakup is receiving serious consideration, it's not clear how it can be done without great risk to the the computer industry--or consumers. "The true horror that Joel Klein must face is that he's winning the case," says Peter Huber, of the Manhattan Institute for Policy Research. "The more you try to refine your solution, the stupider you're going to look."

The last time the government broke up a major company was in 1982, when it worked with a cooperative AT&T to disband the old Bell System. That breakup is generally considered a success. It's less clear, however, whether the disassembling of Microsoft, without the neat geographic and physical divisions that AT&T's operations provided regulators, would create the same viable free-standing offspring.

BETTER IDEA? What's more, it's difficult to craft a solution that works in a market that is rapidly moving beyond the technology--desktop computers--that provided Microsoft with its unique monopoly power. The Web, not PCs, is where the innovation is taking place, which means that Microsoft's Windows monopoly could gradually diminish in importance. For that reason, many Silicon Valley competitors are reluctant to back divestiture. "Breaking them up is not the right answer," says Benchmark Capital partner David Beirne, a key backer of Red Hat Software Inc., which makes a Linux-based operating system that competes with Windows.

So, what's a better idea than breakup? Over the past few months, many of Microsoft's friends and enemies have been feeding remedy ideas to two committees that have been studying the problem--one at the Justice Dept. and another representing the 19 states that joined the federal suit. The members of these committees are the first to admit that there are no easy answers--and that they are a long way from consensus.

Until the trustbusters decide what they want to do, don't expect any quick settlement. Following the Nov. 5 finding of fact, the sides have retired to their corners to prepare briefings on how the law should apply. And the judge should make a final decision by early next year. If he rules against Microsoft, as now appears inevitable, the company could pay a big price: Not only will it face court-imposed remedies, but, now that it has been officially branded a monopoly, it could be the target of civil suits aimed at recovering billions in damages suffered by competitors.

Will Microsoft seek a settlement to avoid all that? Maybe, but, it says, divestiture is not a negotiable option. "We've welcomed any opportunity to settle and we've put a lot of energy into that," CEO William H. Gates III told shareholders at the company's Nov. 10 annual meeting. Microsoft president Steven A. Ballmer, in an interview with BUSINESS WEEK, also insisted that the company would be "happy to resolve this as quickly as possible. But, he added, "We aren't going to compromise on the principle of the freedom to innovate." In other words, the Feds can't tell Microsoft what to put in its products.

Meanwhile, the regulators move forward with discussions about remedies. One state source says that the breakup approach will get detailed consideration. The main argument for divestiture is that Microsoft has so much power, and is so innately aggressive, that nothing else will keep the company from unfairly preying on rivals. Proponents of this approach, including the chief executive of one top computer manufacturer, point to Microsoft's increasingly aggressive marketplace behavior after the 1994 consent decree with regulators on Windows 95 as evidence that rules can't contain the company. "I just don't think they can change their behavior," says this executive. "This is a 20-year-old culture, where the first thing they think about each morning is how to leverage the business. What are they going to do--send all the executives to class, like this was sexual harassment? It won't work. The nation will spend the next 20 years in court trying to enforce it."

If the government tries to break Microsoft up, there are two basic approaches. The first would be to sever the company horizontally, along three major lines of business: Operating systems software, applications programs, and Internet services, sites, and products. The second would be to create three Baby Bills--identical companies that each had the ability to market all of Microsoft's products. Both plans have drawn major criticisms for their severity and awkwardness. Briefly, here is how they work.

-- Horizontal breakup by business units. The theory behind this approach is that it would eliminate any financial incentives the newly created operating systems company would have to favor Microsoft applications or Net products. As a result, it has drawn the support of Netscape Communications founder James Clark and Intuit CEO William Campbell.

But horizontal breakup suffers from a major flaw, the state AGs working on the remedies committees acknowledge: It doesn't alter the situation that gave Microsoft its monopoly power because it doesn't create more competition in desktop computer operating systems. In fact, the new Windows-only company would be likely to use the leverage of its desktop monopoly to quash upstarts, just as Microsoft deflected the threat from Netscape Communications Corp. by bundling a free browser into Windows. "We don't want to have to deal with these same problems all over again," says one member of the state AG's legal team.

That's not the only concern. A horizontal split would also require constant monitoring by Judge Jackson to establish a line between functions that belonged in the operating system and those that would be considered applications. Each time Microsoft wanted to upgrade the operating system, it would have to petition Jackson for the right to do so. That's just the type of excessive regulation that Judge Harold H. Greene indulged in with AT&T. It's also likely to be anathema to Judge Jackson's superiors on the D.C. Circuit Court of Appeals.

-- Baby Bills. Unlike a horizontal divestiture, the Baby Bills approach would give trustbusters what they want: immediate operating systems competition. Like their tongue-in-cheek namesakes, the Baby Bells, the mini-Microsofts would theoretically attack one another with competitive fervor, bust out into new markets, and offer consumers all kinds of great new products while lowering costs.

A NIGHTMARE. That's the theory. The reality is that the Baby Bills could also be problem children. The biggest risk is that the three companies would soon be cranking out incompatible versions of Windows, thereby splintering the standard that has been the foundation of a global PC industry, now approaching 120 million units a year.

Moreover, slicing the company vertically would be a logistical nightmare. First, somebody would have to decide how to split up the employees--a move that would inevitably involve splitting work teams. Regulators would also have to make sure that Microsoft didn't shove all the best people in one company and the losers into the others. One possible approach: allowing Gates to create the divisions himself, but giving him last choice as to which unit he keeps.

There are also financial problems. To make sure the units compete, not collude, employees would only be allowed to own stock in the company they worked for. That means millions of shares and options would have to be liquidated. Other assets would be hard to divide too: All of Microsoft's licensing contracts would be divided up, perhaps harming customers.

-- Auctions. There is another structural remedy, however. This strategy would force the company to auction off the source code underlying its Windows operating system to one or two buyers. These companies would then market Windows clones in competition with Microsoft. In order to help them get off the ground, Gates and his troops would be required to share any of the technological innovations that they had planned for the next several years.

A few months ago, this alternative was put forth by some lawyers for the states and leading economists as the easiest way to cope with Microsoft. Unlike a breakup, it wouldn't involve messy logistics. But as the members of the state and federal remedies teams have investigated this option, they have run into one key problem: None of the companies that were considered potential bidders seems to want to square off against Bill Gates in the operating system business.

Why? Windows has millions of lines of code, and all of the programmers who understand how it all works live in Seattle and "their average salary is $400,000 and they don't want to move," says Michael H. Morris, vice-president and general counsel of Sun Microsystems Inc. Unless these specialists could be lured away from Microsoft, it would take many years before another company was in a position to improve the operating system, design innovative applications, or offer customers quality support. "We wouldn't bid on it," says Morris, whose company had been mentioned by backers of mandatory licensing as a potential buyer. "Why would you as a consumer or an [original equipment manufacturer] buy it from the successful bidder rather than Microsoft? It doesn't make any real-world sense."

Finally, there is the option of opening up the Windows "source code"--publishing the programming interfaces that are needed to make programs Windows-compatible. The code could be sold or even given away--like rival Linux is. That would not create Windows clones. But it would level the playing field for companies that want to compete with Microsoft in Windows applications--a key source of its power.

What should the government do? The bottom line is that there are no perfect options. The trustbusters, paradoxically, may have done too good of a job proving their case. They'll look like wimps if they let Microsoft off too easily. But if they insist on breakup, they could prescribe a cure that's worse than the disease.

    Before it's here, it's on the Bloomberg Terminal. LEARN MORE